diff --git a/.dockerignore b/.dockerignore index d9781fe..755dc76 100644 --- a/.dockerignore +++ b/.dockerignore @@ -1,8 +1,7 @@ -# Secrets — prevent .env files from being baked into the image +# Secrets — prevent .env files and encrypted secrets from being baked into the image .env .env.enc -.env.vault -.env.vault.enc +secrets/ # Version control — .git is huge and not needed in image .git diff --git a/.env.example b/.env.example index d5d801e..7e76ec2 100644 --- a/.env.example +++ b/.env.example @@ -63,6 +63,10 @@ FORGE_BOT_USERNAMES=dev-bot,review-bot,planner-bot,gardener-bot,vault-bot,superv WOODPECKER_TOKEN= # [SECRET] Woodpecker API token WOODPECKER_SERVER=http://localhost:8000 # [CONFIG] Woodpecker server URL WOODPECKER_AGENT_SECRET= # [SECRET] shared secret for server↔agent auth (auto-generated) +# Woodpecker privileged-plugin allowlist — comma-separated image names +# Add plugins/docker (and others) here to allow privileged execution +WOODPECKER_PLUGINS_PRIVILEGED=plugins/docker + # WOODPECKER_REPO_ID — now per-project, set in projects/*.toml [ci] section # Woodpecker Postgres (for direct DB queries) @@ -79,24 +83,42 @@ FORWARD_AUTH_SECRET= # [SECRET] Shared secret for Caddy ↔ # ── Vault-only secrets (DO NOT put these in .env) ──────────────────────── # These tokens grant access to external systems (GitHub, ClawHub, deploy targets). -# They live ONLY in .env.vault.enc and are injected into the ephemeral runner -# container at fire time (#745). lib/env.sh explicitly unsets them so agents -# can never hold them directly — all external actions go through vault dispatch. +# They live ONLY in secrets/.enc (age-encrypted, one file per key) and are +# decrypted into the ephemeral runner container at fire time (#745, #777). +# lib/env.sh explicitly unsets them so agents can never hold them directly — +# all external actions go through vault dispatch. # # GITHUB_TOKEN — GitHub API access (publish, deploy, post) # CLAWHUB_TOKEN — ClawHub registry credentials (publish) +# CADDY_SSH_KEY — SSH key for Caddy log collection # (deploy keys) — SSH keys for deployment targets # -# To manage vault secrets: disinto secrets edit-vault -# (vault redesign in progress: PR-based approval, see #73-#77) +# To manage secrets: disinto secrets add/show/remove/list # ── Project-specific secrets ────────────────────────────────────────────── # Store all project secrets here so formulas reference env vars, never hardcode. BASE_RPC_URL= # [SECRET] on-chain RPC endpoint +# ── Local Qwen dev agent (optional) ────────────────────────────────────── +# Set ENABLE_LLAMA_AGENT=1 to emit agents-llama in docker-compose.yml. +# Requires a running llama-server reachable at ANTHROPIC_BASE_URL. +# See docs/agents-llama.md for details. +ENABLE_LLAMA_AGENT=0 # [CONFIG] 1 = enable agents-llama service +ANTHROPIC_BASE_URL= # [CONFIG] e.g. http://host.docker.internal:8081 + # ── Tuning ──────────────────────────────────────────────────────────────── CLAUDE_TIMEOUT=7200 # [CONFIG] max seconds per Claude invocation +# ── Host paths (Nomad-portable) ──────────────────────────────────────────── +# These env vars externalize host-side bind-mount paths from docker-compose.yml. +# At cutover, Nomad jobspecs reference the same vars — no path translation. +# Defaults point at current paths so an empty .env override still works. +CLAUDE_BIN_DIR=/usr/local/bin/claude # [CONFIG] host path to claude CLI binary (resolved by `disinto init`) +CLAUDE_CONFIG_FILE=${HOME}/.claude.json # [CONFIG] host path to claude config JSON file +CLAUDE_DIR=${HOME}/.claude # [CONFIG] host path to .claude directory (reproduce/edge) +AGENT_SSH_DIR=${HOME}/.ssh # [CONFIG] host path to SSH keys directory +SOPS_AGE_DIR=${HOME}/.config/sops/age # [CONFIG] host path to SOPS age key directory + # ── Claude Code shared OAuth state ───────────────────────────────────────── # Shared directory used by every factory container so Claude Code's internal # proper-lockfile-based OAuth refresh lock works across containers. Both diff --git a/.gitignore b/.gitignore index 2fd9aed..21c6fbc 100644 --- a/.gitignore +++ b/.gitignore @@ -3,7 +3,6 @@ # Encrypted secrets — safe to commit (SOPS-encrypted with age) !.env.enc -!.env.vault.enc !.sops.yaml # Per-box project config (generated by disinto init) @@ -33,6 +32,9 @@ docker/agents/bin/ # Note: This file is now committed to track volume mount configuration # docker-compose.yml +# Generated Caddyfile — single source of truth is generate_caddyfile in lib/generators.sh +docker/Caddyfile + # Python bytecode __pycache__/ *.pyc diff --git a/.woodpecker/ops-filer.yml b/.woodpecker/ops-filer.yml deleted file mode 100644 index 98c5bb2..0000000 --- a/.woodpecker/ops-filer.yml +++ /dev/null @@ -1,36 +0,0 @@ -# .woodpecker/ops-filer.yml — Sub-issue filer pipeline (#764) -# -# Triggered on push to main of the ops repo after a sprint PR merges. -# Parses sprints/*.md for ## Sub-issues blocks and files them on the -# project repo via filer-bot (FORGE_FILER_TOKEN). -# -# NOTE: This pipeline runs on the ops repo. It must be registered in the -# ops repo's Woodpecker project. The filer script (lib/sprint-filer.sh) -# lives in the code repo and is cloned into the workspace. -# -# Idempotency: safe to re-run — each sub-issue carries a decomposed-from -# marker that the filer checks before creating. - -when: - branch: main - event: push - -steps: - - name: file-subissues - image: alpine:3 - commands: - - apk add --no-cache bash curl jq - # Clone the code repo to get the filer script - - AUTH_URL=$(printf '%s' "${FORGE_URL}/disinto-admin/disinto.git" | sed "s|://|://token:${FORGE_FILER_TOKEN}@|") - - git clone --depth 1 "$AUTH_URL" /tmp/code-repo - # Run filer against all sprint files in the ops repo workspace - - bash /tmp/code-repo/lib/sprint-filer.sh --all sprints/ - environment: - FORGE_FILER_TOKEN: - from_secret: forge_filer_token - FORGE_URL: - from_secret: forge_url - FORGE_API: - from_secret: forge_api - FORGE_API_BASE: - from_secret: forge_api_base diff --git a/.woodpecker/publish-images.yml b/.woodpecker/publish-images.yml new file mode 100644 index 0000000..15f373d --- /dev/null +++ b/.woodpecker/publish-images.yml @@ -0,0 +1,64 @@ +# .woodpecker/publish-images.yml — Build and push versioned container images +# Triggered on tag pushes (e.g. v1.2.3). Builds and pushes: +# - ghcr.io/disinto/agents: +# - ghcr.io/disinto/reproduce: +# - ghcr.io/disinto/edge: +# +# Requires GHCR_TOKEN secret configured in Woodpecker with push access +# to ghcr.io/disinto. + +when: + event: tag + ref: refs/tags/v* + +clone: + git: + image: alpine/git + commands: + - AUTH_URL=$(printf '%s' "$CI_REPO_CLONE_URL" | sed "s|://|://token:$FORGE_TOKEN@|") + - git clone --depth 1 "$AUTH_URL" . + - git fetch --depth 1 origin "$CI_COMMIT_REF" + - git checkout FETCH_HEAD + +steps: + - name: build-and-push-agents + image: plugins/docker + settings: + repo: ghcr.io/disinto/agents + registry: ghcr.io + dockerfile: docker/agents/Dockerfile + context: . + tags: + - ${CI_COMMIT_TAG} + - latest + username: disinto + password: + from_secret: GHCR_TOKEN + + - name: build-and-push-reproduce + image: plugins/docker + settings: + repo: ghcr.io/disinto/reproduce + registry: ghcr.io + dockerfile: docker/reproduce/Dockerfile + context: . + tags: + - ${CI_COMMIT_TAG} + - latest + username: disinto + password: + from_secret: GHCR_TOKEN + + - name: build-and-push-edge + image: plugins/docker + settings: + repo: ghcr.io/disinto/edge + registry: ghcr.io + dockerfile: docker/edge/Dockerfile + context: docker/edge + tags: + - ${CI_COMMIT_TAG} + - latest + username: disinto + password: + from_secret: GHCR_TOKEN diff --git a/.woodpecker/run-secret-scan.sh b/.woodpecker/run-secret-scan.sh new file mode 100644 index 0000000..e8d7d5d --- /dev/null +++ b/.woodpecker/run-secret-scan.sh @@ -0,0 +1,68 @@ +#!/usr/bin/env bash +set -euo pipefail +# run-secret-scan.sh — CI wrapper for lib/secret-scan.sh +# +# Scans files changed in this PR for plaintext secrets. +# Exits non-zero if any secret is detected. + +# shellcheck source=../lib/secret-scan.sh +source lib/secret-scan.sh + +# Path patterns considered secret-adjacent +SECRET_PATH_PATTERNS=( + '\.env' + 'tools/vault-.*\.sh' + 'nomad/' + 'vault/' + 'action-vault/' + 'lib/hvault\.sh' + 'lib/action-vault\.sh' +) + +# Build a single regex from patterns +path_regex=$(printf '%s|' "${SECRET_PATH_PATTERNS[@]}") +path_regex="${path_regex%|}" + +# Get files changed in this PR vs target branch. +# Note: shallow clone (depth 50) may lack the merge base for very large PRs, +# causing git diff to fail — || true means the gate skips rather than blocks. +changed_files=$(git diff --name-only --diff-filter=ACMR "origin/${CI_COMMIT_TARGET_BRANCH}...HEAD" || true) + +if [ -z "$changed_files" ]; then + echo "secret-scan: no changed files found, skipping" + exit 0 +fi + +# Filter to secret-adjacent paths only +target_files=$(printf '%s\n' "$changed_files" | grep -E "$path_regex" || true) + +if [ -z "$target_files" ]; then + echo "secret-scan: no secret-adjacent files changed, skipping" + exit 0 +fi + +echo "secret-scan: scanning $(printf '%s\n' "$target_files" | wc -l) file(s):" +printf ' %s\n' "$target_files" + +failures=0 +while IFS= read -r file; do + # Skip deleted files / non-existent + [ -f "$file" ] || continue + # Skip binary files + file -b --mime-encoding "$file" 2>/dev/null | grep -q binary && continue + + content=$(cat "$file") + if ! scan_for_secrets "$content"; then + echo "FAIL: secret detected in $file" + failures=$((failures + 1)) + fi +done <<< "$target_files" + +if [ "$failures" -gt 0 ]; then + echo "" + echo "secret-scan: $failures file(s) contain potential secrets — merge blocked" + echo "If these are false positives, verify patterns in lib/secret-scan.sh" + exit 1 +fi + +echo "secret-scan: all files clean" diff --git a/.woodpecker/secret-scan.yml b/.woodpecker/secret-scan.yml new file mode 100644 index 0000000..7db9c50 --- /dev/null +++ b/.woodpecker/secret-scan.yml @@ -0,0 +1,32 @@ +# .woodpecker/secret-scan.yml — Block PRs that leak plaintext secrets +# +# Triggers on pull requests touching secret-adjacent paths. +# Sources lib/secret-scan.sh and scans each changed file's content. +# Exits non-zero if any potential secret is detected. + +when: + - event: pull_request + path: + - ".env*" + - "tools/vault-*.sh" + - "nomad/**/*" + - "vault/**/*" + - "action-vault/**/*" + - "lib/hvault.sh" + - "lib/action-vault.sh" + +clone: + git: + image: alpine/git + commands: + - AUTH_URL=$(printf '%s' "$CI_REPO_CLONE_URL" | sed "s|://|://token:$FORGE_TOKEN@|") + - git clone --depth 50 "$AUTH_URL" . + - git fetch --depth 50 origin "$CI_COMMIT_REF" "$CI_COMMIT_TARGET_BRANCH" + - git checkout FETCH_HEAD + +steps: + - name: secret-scan + image: alpine:3 + commands: + - apk add --no-cache bash git grep file + - bash .woodpecker/run-secret-scan.sh diff --git a/AGENTS.md b/AGENTS.md index 3a7fc48..c893b09 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -1,4 +1,4 @@ - + # Disinto — Agent Instructions ## What this repo is @@ -31,11 +31,11 @@ disinto/ (code repo) ├── supervisor/ supervisor-run.sh — formula-driven health monitoring (polling-loop executor) │ preflight.sh — pre-flight data collection for supervisor formula ├── architect/ architect-run.sh — strategic decomposition of vision into sprints -├── vault/ vault-env.sh — shared env setup (vault redesign in progress, see #73-#77) +├── action-vault/ vault-env.sh — shared env setup (vault redesign in progress, see #73-#77) │ SCHEMA.md — vault item schema documentation │ validate.sh — vault item validator │ examples/ — example vault action TOMLs (promote, publish, release, webhook-call) -├── lib/ env.sh, agent-sdk.sh, ci-helpers.sh, ci-debug.sh, load-project.sh, parse-deps.sh, guard.sh, mirrors.sh, pr-lifecycle.sh, issue-lifecycle.sh, worktree.sh, formula-session.sh, stack-lock.sh, forge-setup.sh, forge-push.sh, ops-setup.sh, ci-setup.sh, generators.sh, hire-agent.sh, release.sh, build-graph.py, branch-protection.sh, secret-scan.sh, tea-helpers.sh, vault.sh, ci-log-reader.py, git-creds.sh, sprint-filer.sh +├── lib/ env.sh, agent-sdk.sh, ci-helpers.sh, ci-debug.sh, load-project.sh, parse-deps.sh, guard.sh, mirrors.sh, pr-lifecycle.sh, issue-lifecycle.sh, worktree.sh, formula-session.sh, stack-lock.sh, forge-setup.sh, forge-push.sh, ops-setup.sh, ci-setup.sh, generators.sh, hire-agent.sh, release.sh, build-graph.py, branch-protection.sh, secret-scan.sh, tea-helpers.sh, action-vault.sh, ci-log-reader.py, git-creds.sh, sprint-filer.sh, hvault.sh │ hooks/ — Claude Code session hooks (on-compact-reinject, on-idle-stop, on-phase-change, on-pretooluse-guard, on-session-end, on-stop-failure) ├── projects/ *.toml.example — templates; *.toml — local per-box config (gitignored) ├── formulas/ Issue templates (TOML specs for multi-step agent tasks) @@ -43,7 +43,7 @@ disinto/ (code repo) ├── tools/ Operational tools: edge-control/ (register.sh, install.sh, verify-chat-sandbox.sh) ├── docs/ Protocol docs (PHASE-PROTOCOL.md, EVIDENCE-ARCHITECTURE.md) ├── site/ disinto.ai website content -├── tests/ Test files (mock-forgejo.py, smoke-init.sh) +├── tests/ Test files (mock-forgejo.py, smoke-init.sh, lib-hvault.bats) ├── templates/ Issue templates ├── bin/ The `disinto` CLI script ├── disinto-factory/ Setup documentation and skill @@ -86,7 +86,7 @@ Each agent has a `.profile` repository on Forgejo storing `knowledge/lessons-lea - All scripts start with `#!/usr/bin/env bash` and `set -euo pipefail` - Source shared environment: `source "$(dirname "$0")/../lib/env.sh"` - Log to `$LOGFILE` using the `log()` function from env.sh or defined locally -- Never hardcode secrets — agent secrets come from `.env.enc`, vault secrets from `.env.vault.enc` (or `.env`/`.env.vault` fallback) +- Never hardcode secrets — agent secrets come from `.env.enc`, vault secrets from `secrets/.enc` (age-encrypted, one file per key) - Never embed secrets in issue bodies, PR descriptions, or comments — use env var references (e.g. `$BASE_RPC_URL`) - ShellCheck must pass (CI runs `shellcheck` on all `.sh` files) - Avoid duplicate code — shared helpers go in `lib/` @@ -114,10 +114,12 @@ bash dev/phase-test.sh | Planner | `planner/` | Strategic planning | [planner/AGENTS.md](planner/AGENTS.md) | | Predictor | `predictor/` | Infrastructure pattern detection | [predictor/AGENTS.md](predictor/AGENTS.md) | | Architect | `architect/` | Strategic decomposition (read-only on project repo) | [architect/AGENTS.md](architect/AGENTS.md) | -| Filer | `lib/sprint-filer.sh` | Sub-issue filing from merged sprint PRs | `.woodpecker/ops-filer.yml` | +| Filer | `lib/sprint-filer.sh` | Sub-issue filing from merged sprint PRs | ops repo pipeline (deferred, see #779) | | Reproduce | `docker/reproduce/` | Bug reproduction using Playwright MCP | `formulas/reproduce.toml` | | Triage | `docker/reproduce/` | Deep root cause analysis | `formulas/triage.toml` | | Edge dispatcher | `docker/edge/` | Polls ops repo for vault actions, executes via Claude sessions | `docker/edge/dispatcher.sh` | +| agents-llama | `docker/agents/` (same image) | Local-Qwen dev agent (`AGENT_ROLES=dev`), gated on `ENABLE_LLAMA_AGENT=1` | [docs/agents-llama.md](docs/agents-llama.md) | +| agents-llama-all | `docker/agents/` (same image) | Local-Qwen all-roles agent (all 7 roles), profile `agents-llama-all` | [docs/agents-llama.md](docs/agents-llama.md) | > **Vault:** Being redesigned as a PR-based approval workflow (issues #73-#77). > See [docs/VAULT.md](docs/VAULT.md) for the vault PR workflow details. @@ -178,8 +180,8 @@ Humans write these. Agents read and enforce them. | AD-002 | **Concurrency is bounded per LLM backend, not per project.** One concurrent Claude session per OAuth credential pool; one concurrent session per llama-server instance. Containers with disjoint backends may run in parallel. | The single-thread invariant is about *backends*, not pipelines. **(a) Anthropic OAuth credentials race on token refresh** — each container uses a per-session `CLAUDE_CONFIG_DIR`, so Claude Code's native lockfile-based OAuth refresh handles contention automatically without external serialization. (Legacy: set `CLAUDE_EXTERNAL_LOCK=1` to re-enable the old `flock session.lock` wrapper for rollback.) **(b) llama-server has finite VRAM and one KV cache** — parallel inference thrashes the cache and risks OOM. All llama-backed agents serialize on the same lock. **(c) Disjoint backends are free to parallelize.** Today `disinto-agents` (Anthropic OAuth, runs `review,gardener`) runs concurrently with `disinto-agents-llama` (llama, runs `dev`) on the same project — they share neither OAuth state nor llama VRAM. **(d) Per-project work-conflict safety** (no duplicate dev work, no merge conflicts on the same branch) is enforced by `issue_claim` (assignee + `in-progress` label) and per-issue worktrees — that's a separate guard that does NOT depend on this AD. | | AD-003 | The runtime creates and destroys, the formula preserves. | Runtime manages worktrees/sessions/temp. Formulas commit knowledge to git before signaling done. | | AD-004 | Event-driven > polling > fixed delays. | Never `waitForTimeout` or hardcoded sleep. Use phase files, webhooks, or poll loops with backoff. | -| AD-005 | Secrets via env var indirection, never in issue bodies. | Issue bodies become code. Agent secrets go in `.env.enc`, vault secrets in `.env.vault.enc` (SOPS-encrypted when available; plaintext `.env`/`.env.vault` fallback supported). Referenced as `$VAR_NAME`. Runner gets only vault secrets; agents get only agent secrets. | -| AD-006 | External actions go through vault dispatch, never direct. | Agents build addressables; only the vault exercises them (publishes, deploys, posts). Tokens for external systems (`GITHUB_TOKEN`, `CLAWHUB_TOKEN`, deploy keys) live only in `.env.vault.enc` and are injected into the ephemeral runner container. `lib/env.sh` unsets them so agents never hold them. PRs with direct external actions without vault dispatch get REQUEST_CHANGES. (Vault redesign in progress: PR-based approval on ops repo, see #73-#77) | +| AD-005 | Secrets via env var indirection, never in issue bodies. | Issue bodies become code. Agent secrets go in `.env.enc` (SOPS-encrypted), vault secrets in `secrets/.enc` (age-encrypted, one file per key). Referenced as `$VAR_NAME`. Runner gets only vault secrets; agents get only agent secrets. | +| AD-006 | External actions go through vault dispatch, never direct. | Agents build addressables; only the vault exercises them (publishes, deploys, posts). Tokens for external systems (`GITHUB_TOKEN`, `CLAWHUB_TOKEN`, deploy keys) live only in `secrets/.enc` and are decrypted into the ephemeral runner container. `lib/env.sh` unsets them so agents never hold them. PRs with direct external actions without vault dispatch get REQUEST_CHANGES. (Vault redesign in progress: PR-based approval on ops repo, see #73-#77) | **Who enforces what:** - **Gardener** checks open backlog issues against ADs during grooming; closes violations with a comment referencing the AD number. @@ -187,8 +189,6 @@ Humans write these. Agents read and enforce them. - **Dev-agent** reads AGENTS.md before implementing; refuses work that violates ADs. - **AD-002 is a runtime invariant; nothing for the gardener to check at issue-groom time.** OAuth concurrency is handled by per-session `CLAUDE_CONFIG_DIR` isolation (with `CLAUDE_EXTERNAL_LOCK` as a rollback flag). Per-issue work is enforced by `issue_claim`. A violation manifests as a 401 or VRAM OOM in agent logs, not as a malformed issue. ---- - ## Phase-Signaling Protocol When running as a persistent tmux session, Claude must signal the orchestrator @@ -197,5 +197,4 @@ at each phase boundary by writing to a phase file (e.g. Key phases: `PHASE:awaiting_ci` → `PHASE:awaiting_review` → `PHASE:done`. Also: `PHASE:escalate` (needs human input), `PHASE:failed`. - See [docs/PHASE-PROTOCOL.md](docs/PHASE-PROTOCOL.md) for the complete spec, orchestrator reaction matrix, sequence diagram, and crash recovery. diff --git a/vault/SCHEMA.md b/action-vault/SCHEMA.md similarity index 95% rename from vault/SCHEMA.md rename to action-vault/SCHEMA.md index adab177..dd84fb8 100644 --- a/vault/SCHEMA.md +++ b/action-vault/SCHEMA.md @@ -50,7 +50,7 @@ blast_radius = "low" # optional: overrides policy.toml tier ("low"|"medium ## Secret Names -Secret names must be defined in `.env.vault.enc` on the ops repo. The vault validates that requested secrets exist in the allowlist before execution. +Secret names must have a corresponding `secrets/.enc` file (age-encrypted). The vault validates that requested secrets exist in the allowlist before execution. Common secret names: - `CLAWHUB_TOKEN` - Token for ClawHub skill publishing diff --git a/vault/classify.sh b/action-vault/classify.sh similarity index 100% rename from vault/classify.sh rename to action-vault/classify.sh diff --git a/vault/examples/promote.toml b/action-vault/examples/promote.toml similarity index 100% rename from vault/examples/promote.toml rename to action-vault/examples/promote.toml diff --git a/vault/examples/publish.toml b/action-vault/examples/publish.toml similarity index 100% rename from vault/examples/publish.toml rename to action-vault/examples/publish.toml diff --git a/vault/examples/release.toml b/action-vault/examples/release.toml similarity index 100% rename from vault/examples/release.toml rename to action-vault/examples/release.toml diff --git a/vault/examples/webhook-call.toml b/action-vault/examples/webhook-call.toml similarity index 100% rename from vault/examples/webhook-call.toml rename to action-vault/examples/webhook-call.toml diff --git a/vault/policy.toml b/action-vault/policy.toml similarity index 100% rename from vault/policy.toml rename to action-vault/policy.toml diff --git a/vault/validate.sh b/action-vault/validate.sh similarity index 100% rename from vault/validate.sh rename to action-vault/validate.sh diff --git a/vault/vault-env.sh b/action-vault/vault-env.sh similarity index 99% rename from vault/vault-env.sh rename to action-vault/vault-env.sh index 4234774..ec4c83b 100644 --- a/vault/vault-env.sh +++ b/action-vault/vault-env.sh @@ -28,7 +28,7 @@ fi # VAULT ACTION VALIDATION # ============================================================================= -# Allowed secret names - must match keys in .env.vault.enc +# Allowed secret names - must match files in secrets/.enc VAULT_ALLOWED_SECRETS="CLAWHUB_TOKEN GITHUB_TOKEN CODEBERG_TOKEN DEPLOY_KEY NPM_TOKEN DOCKER_HUB_TOKEN" # Allowed mount aliases — well-known file-based credential directories diff --git a/architect/AGENTS.md b/architect/AGENTS.md index e705f23..deee9cf 100644 --- a/architect/AGENTS.md +++ b/architect/AGENTS.md @@ -1,4 +1,4 @@ - + # Architect — Agent Instructions ## What this agent is @@ -96,6 +96,7 @@ Run via `architect/architect-run.sh`, which: - Acquires a poll-loop lock (via `acquire_lock`) and checks available memory - Cleans up per-issue scratch files from previous runs (`/tmp/architect-{project}-scratch-*.md`) - Sources shared libraries (env.sh, formula-session.sh) +- Exports `FORGE_TOKEN_OVERRIDE="${FORGE_ARCHITECT_TOKEN}"` BEFORE sourcing env.sh, ensuring architect-bot identity survives re-sourcing (#762) - Uses FORGE_ARCHITECT_TOKEN for authentication - Processes existing architect PRs via bash-driven design phase - Loads the formula and builds context from VISION.md, AGENTS.md, and ops repo diff --git a/bin/disinto b/bin/disinto index bbb11ec..00404e6 100755 --- a/bin/disinto +++ b/bin/disinto @@ -81,9 +81,12 @@ Init options: --repo-root Local clone path (default: ~/name) --ci-id Woodpecker CI repo ID (default: 0 = no CI) --forge-url Forge base URL (default: http://localhost:3000) + --backend Orchestration backend: docker (default) | nomad (stub, S0.1) --bare Skip compose generation (bare-metal setup) + --build Use local docker build instead of registry images (dev mode) --yes Skip confirmation prompts --rotate-tokens Force regeneration of all bot tokens/passwords (idempotent by default) + --dry-run Print every intended action without executing Hire an agent options: --formula Path to role formula TOML (default: formulas/.toml) @@ -642,6 +645,19 @@ prompt_admin_password() { # ── init command ───────────────────────────────────────────────────────────── +# Nomad backend init — stub for the Nomad+Vault migration (issue #821, S0.1). +# Real implementation lands across S0.2–S0.5. Exists so --backend=nomad fails +# loud instead of silently routing through the docker path. +_disinto_init_nomad() { + local dry_run="${1:-false}" + if [ "$dry_run" = "true" ]; then + echo "nomad backend: stub — will be implemented by S0.2–S0.5" + exit 0 + fi + echo "ERROR: nomad backend not yet implemented (stub)" >&2 + exit 99 +} + disinto_init() { local repo_url="${1:-}" if [ -z "$repo_url" ]; then @@ -652,20 +668,40 @@ disinto_init() { shift # Parse flags - local branch="" repo_root="" ci_id="0" auto_yes=false forge_url_flag="" bare=false rotate_tokens=false + local branch="" repo_root="" ci_id="0" auto_yes=false forge_url_flag="" bare=false rotate_tokens=false use_build=false dry_run=false backend="docker" while [ $# -gt 0 ]; do case "$1" in --branch) branch="$2"; shift 2 ;; --repo-root) repo_root="$2"; shift 2 ;; --ci-id) ci_id="$2"; shift 2 ;; --forge-url) forge_url_flag="$2"; shift 2 ;; + --backend) backend="$2"; shift 2 ;; + --backend=*) backend="${1#--backend=}"; shift ;; --bare) bare=true; shift ;; + --build) use_build=true; shift ;; --yes) auto_yes=true; shift ;; --rotate-tokens) rotate_tokens=true; shift ;; + --dry-run) dry_run=true; shift ;; *) echo "Unknown option: $1" >&2; exit 1 ;; esac done + # Validate backend + case "$backend" in + docker|nomad) ;; + *) echo "Error: invalid --backend value '${backend}' (expected: docker|nomad)" >&2; exit 1 ;; + esac + + # Dispatch on backend — nomad path is a stub for now (issue #821, S0.1). + # Subsequent S0.x issues will replace _disinto_init_nomad with real logic + # without touching flag parsing or this dispatch. + if [ "$backend" = "nomad" ]; then + _disinto_init_nomad "$dry_run" + # shellcheck disable=SC2317 # _disinto_init_nomad always exits today; + # `return` is defensive against future refactors. + return + fi + # Export bare-metal flag for setup_forge export DISINTO_BARE="$bare" @@ -738,12 +774,92 @@ p.write_text(text) fi fi + # ── Dry-run mode: report intended actions and exit ───────────────────────── + if [ "$dry_run" = true ]; then + echo "" + echo "── Dry-run: intended actions ────────────────────────────" + local env_file="${FACTORY_ROOT}/.env" + local rr="${repo_root:-/home/${USER}/${project_name}}" + + if [ "$bare" = false ]; then + [ -f "${FACTORY_ROOT}/docker-compose.yml" ] \ + && echo "[skip] docker-compose.yml (exists)" \ + || echo "[create] docker-compose.yml" + fi + + [ -f "$env_file" ] \ + && echo "[exists] .env" \ + || echo "[create] .env" + + # Report token state from .env + if [ -f "$env_file" ]; then + local _var + for _var in FORGE_ADMIN_TOKEN HUMAN_TOKEN FORGE_TOKEN FORGE_REVIEW_TOKEN \ + FORGE_PLANNER_TOKEN FORGE_GARDENER_TOKEN FORGE_VAULT_TOKEN \ + FORGE_SUPERVISOR_TOKEN FORGE_PREDICTOR_TOKEN FORGE_ARCHITECT_TOKEN; do + if grep -q "^${_var}=" "$env_file" 2>/dev/null; then + echo "[keep] ${_var} (preserved)" + else + echo "[create] ${_var}" + fi + done + else + echo "[create] all tokens and passwords" + fi + + echo "" + echo "[ensure] Forgejo admin user 'disinto-admin'" + echo "[ensure] 8 bot users: dev-bot, review-bot, planner-bot, gardener-bot, vault-bot, supervisor-bot, predictor-bot, architect-bot" + echo "[ensure] 2 llama bot users: dev-qwen, dev-qwen-nightly" + echo "[ensure] .profile repos for all bots" + echo "[ensure] repo ${forge_repo} on Forgejo with collaborators" + echo "[run] preflight checks" + + [ -d "${rr}/.git" ] \ + && echo "[skip] clone ${rr} (exists)" \ + || echo "[clone] ${repo_url} -> ${rr}" + + echo "[push] to local Forgejo" + echo "[ensure] ops repo disinto-admin/${project_name}-ops" + echo "[ensure] branch protection on ${forge_repo}" + + [ "$toml_exists" = true ] \ + && echo "[skip] ${toml_path} (exists)" \ + || echo "[create] ${toml_path}" + + if [ "$bare" = false ]; then + echo "[ensure] Woodpecker OAuth2 app" + echo "[ensure] Chat OAuth2 app" + echo "[ensure] WOODPECKER_AGENT_SECRET in .env" + fi + + echo "[ensure] labels on ${forge_repo}" + + [ -f "${rr}/VISION.md" ] \ + && echo "[skip] VISION.md (exists)" \ + || echo "[create] VISION.md" + + echo "[copy] issue templates" + echo "[ensure] scheduling (cron or compose polling)" + + if [ "$bare" = false ]; then + echo "[start] docker compose stack" + echo "[ensure] Woodpecker token + repo activation" + fi + + echo "[ensure] CLAUDE_CONFIG_DIR" + echo "[ensure] state files (.dev-active, .reviewer-active, .gardener-active)" + echo "" + echo "Dry run complete — no changes made." + exit 0 + fi + # Generate compose files (unless --bare) if [ "$bare" = false ]; then local forge_port forge_port=$(printf '%s' "$forge_url" | sed -E 's|.*:([0-9]+)/?$|\1|') forge_port="${forge_port:-3000}" - generate_compose "$forge_port" + generate_compose "$forge_port" "$use_build" generate_agent_docker generate_caddyfile generate_staging_index @@ -890,6 +1006,19 @@ p.write_text(text) echo "Config: CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=1 saved to .env" fi + # Write local-Qwen dev agent env keys with safe defaults (#769) + if ! grep -q '^ENABLE_LLAMA_AGENT=' "$env_file" 2>/dev/null; then + cat >> "$env_file" <<'LLAMAENVEOF' + +# Local Qwen dev agent (optional) — set to 1 to enable +ENABLE_LLAMA_AGENT=0 +FORGE_TOKEN_LLAMA= +FORGE_PASS_LLAMA= +ANTHROPIC_BASE_URL= +LLAMAENVEOF + echo "Config: ENABLE_LLAMA_AGENT keys written to .env (disabled by default)" + fi + # Create labels on remote create_labels "$forge_repo" "$forge_url" @@ -1118,8 +1247,6 @@ disinto_secrets() { local subcmd="${1:-}" local enc_file="${FACTORY_ROOT}/.env.enc" local env_file="${FACTORY_ROOT}/.env" - local vault_enc_file="${FACTORY_ROOT}/.env.vault.enc" - local vault_env_file="${FACTORY_ROOT}/.env.vault" # Shared helper: ensure sops+age and .sops.yaml exist _secrets_ensure_sops() { @@ -1165,30 +1292,51 @@ disinto_secrets() { case "$subcmd" in add) - local name="${2:-}" + # Parse flags + local force=false + shift # consume 'add' + while [ $# -gt 0 ]; do + case "$1" in + -f|--force) force=true; shift ;; + -*) echo "Unknown flag: $1" >&2; exit 1 ;; + *) break ;; + esac + done + local name="${1:-}" if [ -z "$name" ]; then - echo "Usage: disinto secrets add " >&2 + echo "Usage: disinto secrets add [-f|--force] " >&2 exit 1 fi _secrets_ensure_age_key mkdir -p "$secrets_dir" - printf 'Enter value for %s: ' "$name" >&2 local value - IFS= read -rs value - echo >&2 + if [ -t 0 ]; then + # Interactive TTY — prompt with hidden input (original behavior) + printf 'Enter value for %s: ' "$name" >&2 + IFS= read -rs value + echo >&2 + else + # Piped/redirected stdin — read raw bytes verbatim + IFS= read -r -d '' value || true + fi if [ -z "$value" ]; then echo "Error: empty value" >&2 exit 1 fi local enc_path="${secrets_dir}/${name}.enc" - if [ -f "$enc_path" ]; then - printf 'Secret %s already exists. Overwrite? [y/N] ' "$name" >&2 - local confirm - read -r confirm - if [ "$confirm" != "y" ] && [ "$confirm" != "Y" ]; then - echo "Aborted." >&2 + if [ -f "$enc_path" ] && [ "$force" = false ]; then + if [ -t 0 ]; then + printf 'Secret %s already exists. Overwrite? [y/N] ' "$name" >&2 + local confirm + read -r confirm + if [ "$confirm" != "y" ] && [ "$confirm" != "Y" ]; then + echo "Aborted." >&2 + exit 1 + fi + else + echo "Error: secret ${name} already exists (use -f to overwrite)" >&2 exit 1 fi fi @@ -1221,6 +1369,37 @@ disinto_secrets() { sops -d "$enc_file" fi ;; + remove) + local name="${2:-}" + if [ -z "$name" ]; then + echo "Usage: disinto secrets remove " >&2 + exit 1 + fi + local enc_path="${secrets_dir}/${name}.enc" + if [ ! -f "$enc_path" ]; then + echo "Error: ${enc_path} not found" >&2 + exit 1 + fi + rm -f "$enc_path" + echo "Removed: ${enc_path}" + ;; + list) + if [ ! -d "$secrets_dir" ]; then + echo "No secrets directory found." >&2 + exit 0 + fi + local found=false + for enc_file_path in "${secrets_dir}"/*.enc; do + [ -f "$enc_file_path" ] || continue + found=true + local secret_name + secret_name=$(basename "$enc_file_path" .enc) + echo "$secret_name" + done + if [ "$found" = false ]; then + echo "No secrets stored." >&2 + fi + ;; edit) if [ ! -f "$enc_file" ]; then echo "Error: ${enc_file} not found. Run 'disinto secrets migrate' first." >&2 @@ -1244,54 +1423,100 @@ disinto_secrets() { rm -f "$env_file" echo "Migrated: .env -> .env.enc (plaintext removed)" ;; - edit-vault) - if [ ! -f "$vault_enc_file" ]; then - echo "Error: ${vault_enc_file} not found. Run 'disinto secrets migrate-vault' first." >&2 + migrate-from-vault) + # One-shot migration: split .env.vault.enc into secrets/.enc files (#777) + local vault_enc_file="${FACTORY_ROOT}/.env.vault.enc" + local vault_env_file="${FACTORY_ROOT}/.env.vault" + local source_file="" + + if [ -f "$vault_enc_file" ] && command -v sops &>/dev/null; then + source_file="$vault_enc_file" + elif [ -f "$vault_env_file" ]; then + source_file="$vault_env_file" + else + echo "Error: neither .env.vault.enc nor .env.vault found — nothing to migrate." >&2 exit 1 fi - sops "$vault_enc_file" - ;; - show-vault) - if [ ! -f "$vault_enc_file" ]; then - echo "Error: ${vault_enc_file} not found." >&2 + + _secrets_ensure_age_key + mkdir -p "$secrets_dir" + + # Decrypt vault to temp dotenv + local tmp_dotenv + tmp_dotenv=$(mktemp /tmp/disinto-vault-migrate-XXXXXX) + trap 'rm -f "$tmp_dotenv"' RETURN + + if [ "$source_file" = "$vault_enc_file" ]; then + if ! sops -d --output-type dotenv "$vault_enc_file" > "$tmp_dotenv" 2>/dev/null; then + rm -f "$tmp_dotenv" + echo "Error: failed to decrypt .env.vault.enc" >&2 + exit 1 + fi + else + cp "$vault_env_file" "$tmp_dotenv" + fi + + # Parse each KEY=VALUE and encrypt into secrets/.enc + local count=0 + local failed=0 + while IFS='=' read -r key value; do + # Skip empty lines and comments + [[ -z "$key" || "$key" =~ ^[[:space:]]*# ]] && continue + # Trim whitespace from key + key=$(echo "$key" | xargs) + [ -z "$key" ] && continue + + local enc_path="${secrets_dir}/${key}.enc" + if printf '%s' "$value" | age -r "$AGE_PUBLIC_KEY" -o "$enc_path" 2>/dev/null; then + # Verify round-trip + local check + check=$(age -d -i "$age_key_file" "$enc_path" 2>/dev/null) || { failed=$((failed + 1)); echo " FAIL (verify): ${key}" >&2; continue; } + if [ "$check" = "$value" ]; then + echo " OK: ${key} -> secrets/${key}.enc" + count=$((count + 1)) + else + echo " FAIL (mismatch): ${key}" >&2 + failed=$((failed + 1)) + fi + else + echo " FAIL (encrypt): ${key}" >&2 + failed=$((failed + 1)) + fi + done < "$tmp_dotenv" + + rm -f "$tmp_dotenv" + + if [ "$failed" -gt 0 ]; then + echo "Error: ${failed} secret(s) failed migration. Vault files NOT removed." >&2 exit 1 fi - sops -d "$vault_enc_file" - ;; - migrate-vault) - if [ ! -f "$vault_env_file" ]; then - echo "Error: ${vault_env_file} not found — nothing to migrate." >&2 - echo " Create .env.vault with vault secrets (GITHUB_TOKEN, deploy keys, etc.)" >&2 - exit 1 + + if [ "$count" -eq 0 ]; then + echo "Warning: no secrets found in vault file." >&2 + else + echo "Migrated ${count} secret(s) to secrets/*.enc" + # Remove old vault files on success + rm -f "$vault_enc_file" "$vault_env_file" + echo "Removed: .env.vault.enc / .env.vault" fi - _secrets_ensure_sops - encrypt_env_file "$vault_env_file" "$vault_enc_file" - # Verify decryption works before removing plaintext - if ! sops -d "$vault_enc_file" >/dev/null 2>&1; then - echo "Error: failed to verify .env.vault.enc decryption" >&2 - rm -f "$vault_enc_file" - exit 1 - fi - rm -f "$vault_env_file" - echo "Migrated: .env.vault -> .env.vault.enc (plaintext removed)" ;; *) cat <&2 Usage: disinto secrets -Individual secrets (secrets/.enc): - add Prompt for value, encrypt, store in secrets/.enc - show Decrypt and print an individual secret +Secrets (secrets/.enc — age-encrypted, one file per key): + add Prompt for value, encrypt, store in secrets/.enc + show Decrypt and print a secret + remove Remove a secret + list List all stored secrets -Agent secrets (.env.enc): - edit Edit agent secrets (FORGE_TOKEN, CLAUDE_API_KEY, etc.) - show Show decrypted agent secrets (no argument) - migrate Encrypt .env -> .env.enc +Agent secrets (.env.enc — sops-encrypted dotenv): + edit Edit agent secrets (FORGE_TOKEN, CLAUDE_API_KEY, etc.) + show Show decrypted agent secrets (no argument) + migrate Encrypt .env -> .env.enc -Vault secrets (.env.vault.enc): - edit-vault Edit vault secrets (GITHUB_TOKEN, deploy keys, etc.) - show-vault Show decrypted vault secrets - migrate-vault Encrypt .env.vault -> .env.vault.enc +Migration: + migrate-from-vault Split .env.vault.enc into secrets/.enc (one-shot) EOF exit 1 ;; @@ -1303,7 +1528,8 @@ EOF disinto_run() { local action_id="${1:?Usage: disinto run }" local compose_file="${FACTORY_ROOT}/docker-compose.yml" - local vault_enc="${FACTORY_ROOT}/.env.vault.enc" + local secrets_dir="${FACTORY_ROOT}/secrets" + local age_key_file="${HOME}/.config/sops/age/keys.txt" if [ ! -f "$compose_file" ]; then echo "Error: docker-compose.yml not found" >&2 @@ -1311,29 +1537,42 @@ disinto_run() { exit 1 fi - if [ ! -f "$vault_enc" ]; then - echo "Error: .env.vault.enc not found — create vault secrets first" >&2 - echo " Run 'disinto secrets migrate-vault' after creating .env.vault" >&2 + if [ ! -d "$secrets_dir" ]; then + echo "Error: secrets/ directory not found — create secrets first" >&2 + echo " Run 'disinto secrets add ' to add secrets" >&2 exit 1 fi - if ! command -v sops &>/dev/null; then - echo "Error: sops not found — required to decrypt vault secrets" >&2 + if ! command -v age &>/dev/null; then + echo "Error: age not found — required to decrypt secrets" >&2 exit 1 fi - # Decrypt vault secrets to temp file + if [ ! -f "$age_key_file" ]; then + echo "Error: age key not found at ${age_key_file}" >&2 + exit 1 + fi + + # Decrypt all secrets/*.enc into a temp env file for the runner local tmp_env - tmp_env=$(mktemp /tmp/disinto-vault-XXXXXX) + tmp_env=$(mktemp /tmp/disinto-secrets-XXXXXX) trap 'rm -f "$tmp_env"' EXIT - if ! sops -d --output-type dotenv "$vault_enc" > "$tmp_env" 2>/dev/null; then - rm -f "$tmp_env" - echo "Error: failed to decrypt .env.vault.enc" >&2 - exit 1 - fi + local count=0 + for enc_path in "${secrets_dir}"/*.enc; do + [ -f "$enc_path" ] || continue + local key + key=$(basename "$enc_path" .enc) + local val + val=$(age -d -i "$age_key_file" "$enc_path" 2>/dev/null) || { + echo "Warning: failed to decrypt ${enc_path}" >&2 + continue + } + printf '%s=%s\n' "$key" "$val" >> "$tmp_env" + count=$((count + 1)) + done - echo "Vault secrets decrypted to tmpfile" + echo "Decrypted ${count} secret(s) to tmpfile" # Run action in ephemeral runner container local rc=0 @@ -1404,21 +1643,96 @@ download_agent_binaries() { # ── up command ──────────────────────────────────────────────────────────────── +# Regenerate a file idempotently: run the generator, compare output, backup if changed. +# Usage: _regen_file [args...] +_regen_file() { + local target="$1"; shift + local generator="$1"; shift + local basename + basename=$(basename "$target") + + # Move existing file aside so the generator (which skips if file exists) + # produces a fresh copy. + local stashed="" + if [ -f "$target" ]; then + stashed=$(mktemp "${target}.stash.XXXXXX") + mv "$target" "$stashed" + fi + + # Run the generator — it writes $target from scratch. + # If the generator fails, restore the stashed original so it is not stranded. + if ! "$generator" "$@"; then + if [ -n "$stashed" ]; then + mv "$stashed" "$target" + fi + return 1 + fi + + if [ -z "$stashed" ]; then + # No previous file — first generation + echo "regenerated: ${basename} (new)" + return + fi + + if cmp -s "$stashed" "$target"; then + # Content unchanged — restore original to preserve mtime + mv "$stashed" "$target" + echo "unchanged: ${basename}" + else + # Content changed — keep new, save old as .prev + mv "$stashed" "${target}.prev" + echo "regenerated: ${basename} (previous saved as ${basename}.prev)" + fi +} + disinto_up() { local compose_file="${FACTORY_ROOT}/docker-compose.yml" + local caddyfile="${FACTORY_ROOT}/docker/Caddyfile" if [ ! -f "$compose_file" ]; then echo "Error: docker-compose.yml not found" >&2 echo " Run 'disinto init ' first (without --bare)" >&2 exit 1 fi - # Pre-build: download binaries to docker/agents/bin/ to avoid network calls during docker build - echo "── Pre-build: downloading agent binaries ────────────────────────" - if ! download_agent_binaries; then - echo "Error: failed to download agent binaries" >&2 - exit 1 + # Parse --no-regen flag; remaining args pass through to docker compose + local no_regen=false + local -a compose_args=() + for arg in "$@"; do + case "$arg" in + --no-regen) no_regen=true ;; + *) compose_args+=("$arg") ;; + esac + done + + # ── Regenerate compose & Caddyfile from generators ────────────────────── + if [ "$no_regen" = true ]; then + echo "Warning: running with unmanaged compose — hand-edits will drift" >&2 + else + # Determine forge_port from FORGE_URL (same logic as init) + local forge_url="${FORGE_URL:-http://localhost:3000}" + local forge_port + forge_port=$(printf '%s' "$forge_url" | sed -E 's|.*:([0-9]+)/?$|\1|') + forge_port="${forge_port:-3000}" + + # Detect build mode from existing compose + local use_build=false + if grep -q '^\s*build:' "$compose_file"; then + use_build=true + fi + + _regen_file "$compose_file" generate_compose "$forge_port" "$use_build" + _regen_file "$caddyfile" generate_caddyfile + fi + + # Pre-build: download binaries only when compose uses local build + if grep -q '^\s*build:' "$compose_file"; then + echo "── Pre-build: downloading agent binaries ────────────────────────" + if ! download_agent_binaries; then + echo "Error: failed to download agent binaries" >&2 + exit 1 + fi + echo "" fi - echo "" # Decrypt secrets to temp .env if SOPS available and .env.enc exists local tmp_env="" @@ -1431,7 +1745,7 @@ disinto_up() { echo "Decrypted secrets for compose" fi - docker compose -f "$compose_file" up -d "$@" + docker compose -f "$compose_file" up -d --build --remove-orphans ${compose_args[@]+"${compose_args[@]}"} echo "Stack is up" # Clean up temp .env (also handled by EXIT trap if compose fails) diff --git a/dev/AGENTS.md b/dev/AGENTS.md index abeb619..4148f46 100644 --- a/dev/AGENTS.md +++ b/dev/AGENTS.md @@ -1,4 +1,4 @@ - + # Dev Agent **Role**: Implement issues autonomously — write code, push branches, address diff --git a/docker-compose.yml b/docker-compose.yml index 65a7f58..ba8c77c 100644 --- a/docker-compose.yml +++ b/docker-compose.yml @@ -14,10 +14,10 @@ services: - agent-data:/home/agent/data - project-repos:/home/agent/repos - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} - - ${HOME}/.claude.json:/home/agent/.claude.json:ro - - CLAUDE_BIN_PLACEHOLDER:/usr/local/bin/claude:ro - - ${HOME}/.ssh:/home/agent/.ssh:ro - - ${HOME}/.config/sops/age:/home/agent/.config/sops/age:ro + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro + - ${SOPS_AGE_DIR:-${HOME}/.config/sops/age}:/home/agent/.config/sops/age:ro - woodpecker-data:/woodpecker-data:ro environment: - FORGE_URL=http://forgejo:3000 @@ -49,6 +49,13 @@ services: - GARDENER_INTERVAL=${GARDENER_INTERVAL:-21600} - ARCHITECT_INTERVAL=${ARCHITECT_INTERVAL:-21600} - PLANNER_INTERVAL=${PLANNER_INTERVAL:-43200} + - SUPERVISOR_INTERVAL=${SUPERVISOR_INTERVAL:-1200} + healthcheck: + test: ["CMD", "pgrep", "-f", "entrypoint.sh"] + interval: 60s + timeout: 5s + retries: 3 + start_period: 30s depends_on: forgejo: condition: service_healthy @@ -70,10 +77,10 @@ services: - agent-data:/home/agent/data - project-repos:/home/agent/repos - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} - - ${HOME}/.claude.json:/home/agent/.claude.json:ro - - CLAUDE_BIN_PLACEHOLDER:/usr/local/bin/claude:ro - - ${HOME}/.ssh:/home/agent/.ssh:ro - - ${HOME}/.config/sops/age:/home/agent/.config/sops/age:ro + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro + - ${SOPS_AGE_DIR:-${HOME}/.config/sops/age}:/home/agent/.config/sops/age:ro - woodpecker-data:/woodpecker-data:ro environment: - FORGE_URL=http://forgejo:3000 @@ -103,6 +110,80 @@ services: - CLAUDE_CONFIG_DIR=${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config} - POLL_INTERVAL=${POLL_INTERVAL:-300} - AGENT_ROLES=dev + healthcheck: + test: ["CMD", "pgrep", "-f", "entrypoint.sh"] + interval: 60s + timeout: 5s + retries: 3 + start_period: 30s + depends_on: + forgejo: + condition: service_healthy + woodpecker: + condition: service_started + networks: + - disinto-net + + agents-llama-all: + build: + context: . + dockerfile: docker/agents/Dockerfile + image: disinto/agents-llama:latest + container_name: disinto-agents-llama-all + restart: unless-stopped + profiles: ["agents-llama-all"] + security_opt: + - apparmor=unconfined + volumes: + - agent-data:/home/agent/data + - project-repos:/home/agent/repos + - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro + - ${SOPS_AGE_DIR:-${HOME}/.config/sops/age}:/home/agent/.config/sops/age:ro + - woodpecker-data:/woodpecker-data:ro + environment: + - FORGE_URL=http://forgejo:3000 + - FORGE_REPO=${FORGE_REPO:-disinto-admin/disinto} + - FORGE_TOKEN=${FORGE_TOKEN_LLAMA:-} + - FORGE_PASS=${FORGE_PASS_LLAMA:-} + - FORGE_REVIEW_TOKEN=${FORGE_REVIEW_TOKEN:-} + - FORGE_PLANNER_TOKEN=${FORGE_PLANNER_TOKEN:-} + - FORGE_GARDENER_TOKEN=${FORGE_GARDENER_TOKEN:-} + - FORGE_VAULT_TOKEN=${FORGE_VAULT_TOKEN:-} + - FORGE_SUPERVISOR_TOKEN=${FORGE_SUPERVISOR_TOKEN:-} + - FORGE_PREDICTOR_TOKEN=${FORGE_PREDICTOR_TOKEN:-} + - FORGE_ARCHITECT_TOKEN=${FORGE_ARCHITECT_TOKEN:-} + - FORGE_FILER_TOKEN=${FORGE_FILER_TOKEN:-} + - FORGE_BOT_USERNAMES=${FORGE_BOT_USERNAMES:-} + - WOODPECKER_TOKEN=${WOODPECKER_TOKEN:-} + - CLAUDE_TIMEOUT=${CLAUDE_TIMEOUT:-7200} + - CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=${CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC:-1} + - CLAUDE_AUTOCOMPACT_PCT_OVERRIDE=60 + - CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1 + - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-} + - ANTHROPIC_BASE_URL=${ANTHROPIC_BASE_URL:-} + - FORGE_ADMIN_PASS=${FORGE_ADMIN_PASS:-} + - DISINTO_CONTAINER=1 + - PROJECT_TOML=projects/disinto.toml + - PROJECT_NAME=${PROJECT_NAME:-project} + - PROJECT_REPO_ROOT=/home/agent/repos/${PROJECT_NAME:-project} + - WOODPECKER_DATA_DIR=/woodpecker-data + - WOODPECKER_REPO_ID=${WOODPECKER_REPO_ID:-} + - CLAUDE_CONFIG_DIR=${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config} + - POLL_INTERVAL=${POLL_INTERVAL:-300} + - GARDENER_INTERVAL=${GARDENER_INTERVAL:-21600} + - ARCHITECT_INTERVAL=${ARCHITECT_INTERVAL:-21600} + - PLANNER_INTERVAL=${PLANNER_INTERVAL:-43200} + - SUPERVISOR_INTERVAL=${SUPERVISOR_INTERVAL:-1200} + - AGENT_ROLES=review,dev,gardener,architect,planner,predictor,supervisor + healthcheck: + test: ["CMD", "pgrep", "-f", "entrypoint.sh"] + interval: 60s + timeout: 5s + retries: 3 + start_period: 30s depends_on: forgejo: condition: service_healthy @@ -122,9 +203,9 @@ services: - /var/run/docker.sock:/var/run/docker.sock - agent-data:/home/agent/data - project-repos:/home/agent/repos - - ${HOME}/.claude:/home/agent/.claude - - /usr/local/bin/claude:/usr/local/bin/claude:ro - - ${HOME}/.ssh:/home/agent/.ssh:ro + - ${CLAUDE_DIR:-${HOME}/.claude}:/home/agent/.claude + - ${CLAUDE_BIN_DIR:-/usr/local/bin/claude}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro env_file: - .env @@ -138,9 +219,9 @@ services: - apparmor=unconfined volumes: - /var/run/docker.sock:/var/run/docker.sock - - /usr/local/bin/claude:/usr/local/bin/claude:ro - - ${HOME}/.claude.json:/root/.claude.json:ro - - ${HOME}/.claude:/root/.claude:ro + - ${CLAUDE_BIN_DIR:-/usr/local/bin/claude}:/usr/local/bin/claude:ro + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/root/.claude.json:ro + - ${CLAUDE_DIR:-${HOME}/.claude}:/root/.claude:ro - disinto-logs:/opt/disinto-logs environment: - FORGE_SUPERVISOR_TOKEN=${FORGE_SUPERVISOR_TOKEN:-} @@ -156,6 +237,12 @@ services: ports: - "80:80" - "443:443" + healthcheck: + test: ["CMD", "curl", "-fsS", "http://localhost:2019/config/"] + interval: 30s + timeout: 5s + retries: 3 + start_period: 15s depends_on: - forgejo networks: diff --git a/docker/agents/Dockerfile b/docker/agents/Dockerfile index 78fbbf6..2939230 100644 --- a/docker/agents/Dockerfile +++ b/docker/agents/Dockerfile @@ -28,6 +28,9 @@ RUN chmod +x /entrypoint.sh # Entrypoint runs polling loop directly, dropping to agent user via gosu. # All scripts execute as the agent user (UID 1000) while preserving env vars. +VOLUME /home/agent/data +VOLUME /home/agent/repos + WORKDIR /home/agent/disinto ENTRYPOINT ["/entrypoint.sh"] diff --git a/docker/agents/entrypoint.sh b/docker/agents/entrypoint.sh index 9df6d01..b7593a2 100644 --- a/docker/agents/entrypoint.sh +++ b/docker/agents/entrypoint.sh @@ -7,14 +7,15 @@ set -euo pipefail # poll scripts. All Docker Compose env vars are inherited (PATH, FORGE_TOKEN, # ANTHROPIC_API_KEY, etc.). # -# AGENT_ROLES env var controls which scripts run: "review,dev,gardener,architect,planner,predictor" -# (default: all six). Uses while-true loop with staggered intervals: +# AGENT_ROLES env var controls which scripts run: "review,dev,gardener,architect,planner,predictor,supervisor" +# (default: all seven). Uses while-true loop with staggered intervals: # - review-poll: every 5 minutes (offset by 0s) # - dev-poll: every 5 minutes (offset by 2 minutes) # - gardener: every GARDENER_INTERVAL seconds (default: 21600 = 6 hours) # - architect: every ARCHITECT_INTERVAL seconds (default: 21600 = 6 hours) # - planner: every PLANNER_INTERVAL seconds (default: 43200 = 12 hours) # - predictor: every 24 hours (288 iterations * 5 min) +# - supervisor: every SUPERVISOR_INTERVAL seconds (default: 1200 = 20 min) DISINTO_BAKED="/home/agent/disinto" DISINTO_LIVE="/home/agent/repos/_factory" @@ -328,7 +329,7 @@ init_state_dir # Parse AGENT_ROLES env var (default: all agents) # Expected format: comma-separated list like "review,dev,gardener" -AGENT_ROLES="${AGENT_ROLES:-review,dev,gardener,architect,planner,predictor}" +AGENT_ROLES="${AGENT_ROLES:-review,dev,gardener,architect,planner,predictor,supervisor}" log "Agent roles configured: ${AGENT_ROLES}" # Poll interval in seconds (5 minutes default) @@ -338,9 +339,10 @@ POLL_INTERVAL="${POLL_INTERVAL:-300}" GARDENER_INTERVAL="${GARDENER_INTERVAL:-21600}" ARCHITECT_INTERVAL="${ARCHITECT_INTERVAL:-21600}" PLANNER_INTERVAL="${PLANNER_INTERVAL:-43200}" +SUPERVISOR_INTERVAL="${SUPERVISOR_INTERVAL:-1200}" log "Entering polling loop (interval: ${POLL_INTERVAL}s, roles: ${AGENT_ROLES})" -log "Gardener interval: ${GARDENER_INTERVAL}s, Architect interval: ${ARCHITECT_INTERVAL}s, Planner interval: ${PLANNER_INTERVAL}s" +log "Gardener interval: ${GARDENER_INTERVAL}s, Architect interval: ${ARCHITECT_INTERVAL}s, Planner interval: ${PLANNER_INTERVAL}s, Supervisor interval: ${SUPERVISOR_INTERVAL}s" # Main polling loop using iteration counter for gardener scheduling iteration=0 @@ -463,6 +465,19 @@ print(cfg.get('primary_branch', 'main')) fi fi fi + + # Supervisor (interval configurable via SUPERVISOR_INTERVAL env var, default 20 min) + if [[ ",${AGENT_ROLES}," == *",supervisor,"* ]]; then + supervisor_iteration=$((iteration * POLL_INTERVAL)) + if [ $((supervisor_iteration % SUPERVISOR_INTERVAL)) -eq 0 ] && [ "$now" -ge "$supervisor_iteration" ]; then + if ! pgrep -f "supervisor-run.sh" >/dev/null; then + log "Running supervisor (iteration ${iteration}, ${SUPERVISOR_INTERVAL}s interval) for ${toml}" + gosu agent bash -c "cd ${DISINTO_DIR} && bash supervisor/supervisor-run.sh \"${toml}\"" >> "${DISINTO_LOG_DIR}/supervisor.log" 2>&1 & + else + log "Skipping supervisor — already running" + fi + fi + fi done sleep "${POLL_INTERVAL}" diff --git a/docker/chat/Dockerfile b/docker/chat/Dockerfile index 81aebbe..3d89863 100644 --- a/docker/chat/Dockerfile +++ b/docker/chat/Dockerfile @@ -30,6 +30,6 @@ WORKDIR /var/chat EXPOSE 8080 HEALTHCHECK --interval=30s --timeout=5s --start-period=10s --retries=3 \ - CMD python3 -c "import urllib.request; urllib.request.urlopen('http://localhost:8080/')" || exit 1 + CMD python3 -c "import urllib.request; urllib.request.urlopen('http://localhost:8080/health')" || exit 1 ENTRYPOINT ["/entrypoint-chat.sh"] diff --git a/docker/chat/server.py b/docker/chat/server.py index ad8897d..6748354 100644 --- a/docker/chat/server.py +++ b/docker/chat/server.py @@ -481,6 +481,14 @@ class ChatHandler(BaseHTTPRequestHandler): parsed = urlparse(self.path) path = parsed.path + # Health endpoint (no auth required) — used by Docker healthcheck + if path == "/health": + self.send_response(200) + self.send_header("Content-Type", "text/plain") + self.end_headers() + self.wfile.write(b"ok\n") + return + # Verify endpoint for Caddy forward_auth (#709) if path == "/chat/auth/verify": self.handle_auth_verify() diff --git a/docker/edge/Dockerfile b/docker/edge/Dockerfile index 6706852..eca7d7e 100644 --- a/docker/edge/Dockerfile +++ b/docker/edge/Dockerfile @@ -1,4 +1,7 @@ FROM caddy:latest RUN apk add --no-cache bash jq curl git docker-cli python3 openssh-client autossh COPY entrypoint-edge.sh /usr/local/bin/entrypoint-edge.sh + +VOLUME /data + ENTRYPOINT ["bash", "/usr/local/bin/entrypoint-edge.sh"] diff --git a/docker/edge/dispatcher.sh b/docker/edge/dispatcher.sh index 67a1ba9..a48abf2 100755 --- a/docker/edge/dispatcher.sh +++ b/docker/edge/dispatcher.sh @@ -8,8 +8,8 @@ # 2. Scan vault/actions/ for TOML files without .result.json # 3. Verify TOML arrived via merged PR with admin merger (Forgejo API) # 4. Validate TOML using vault-env.sh validator -# 5. Decrypt .env.vault.enc and extract only declared secrets -# 6. Launch: docker run --rm disinto/agents:latest +# 5. Decrypt declared secrets via load_secret (lib/env.sh) +# 6. Launch: delegate to _launch_runner_{docker,nomad} backend # 7. Write .result.json with exit code, timestamp, logs summary # # Part of #76. @@ -19,7 +19,7 @@ set -euo pipefail # Resolve script root (parent of lib/) SCRIPT_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)" -# Source shared environment +# Source shared environment (provides load_secret, log helpers, etc.) source "${SCRIPT_ROOT}/../lib/env.sh" # Project TOML location: prefer mounted path, fall back to cloned path @@ -27,26 +27,18 @@ source "${SCRIPT_ROOT}/../lib/env.sh" # the shallow clone only has .toml.example files. PROJECTS_DIR="${PROJECTS_DIR:-${FACTORY_ROOT:-/opt/disinto}-projects}" -# Load vault secrets after env.sh (env.sh unsets them for agent security) -# Vault secrets must be available to the dispatcher -if [ -f "$FACTORY_ROOT/.env.vault.enc" ] && command -v sops &>/dev/null; then - set -a - eval "$(sops -d --output-type dotenv "$FACTORY_ROOT/.env.vault.enc" 2>/dev/null)" \ - || echo "Warning: failed to decrypt .env.vault.enc — vault secrets not loaded" >&2 - set +a -elif [ -f "$FACTORY_ROOT/.env.vault" ]; then - set -a - # shellcheck source=/dev/null - source "$FACTORY_ROOT/.env.vault" - set +a -fi +# ----------------------------------------------------------------------------- +# Backend selection: DISPATCHER_BACKEND={docker,nomad} +# Default: docker. nomad lands as a pure addition during migration Step 5. +# ----------------------------------------------------------------------------- +DISPATCHER_BACKEND="${DISPATCHER_BACKEND:-docker}" # Ops repo location (vault/actions directory) OPS_REPO_ROOT="${OPS_REPO_ROOT:-/home/debian/disinto-ops}" VAULT_ACTIONS_DIR="${OPS_REPO_ROOT}/vault/actions" # Vault action validation -VAULT_ENV="${SCRIPT_ROOT}/../vault/vault-env.sh" +VAULT_ENV="${SCRIPT_ROOT}/../action-vault/vault-env.sh" # Admin users who can merge vault PRs (from issue #77) # Comma-separated list of Forgejo usernames with admin role @@ -350,73 +342,113 @@ get_dispatch_mode() { fi } -# Write result file for an action -# Usage: write_result -write_result() { +# Commit result.json to the ops repo via git push (portable, no bind-mount). +# +# Clones the ops repo into a scratch directory, writes the result file, +# commits as vault-bot, and pushes to the primary branch. +# Idempotent: skips if result.json already exists upstream. +# Retries on push conflict with rebase-and-push (handles concurrent merges). +# +# Usage: commit_result_via_git +commit_result_via_git() { local action_id="$1" local exit_code="$2" local logs="$3" - local result_file="${VAULT_ACTIONS_DIR}/${action_id}.result.json" + local result_relpath="vault/actions/${action_id}.result.json" + local ops_clone_url="${FORGE_URL}/${FORGE_OPS_REPO}.git" + local branch="${PRIMARY_BRANCH:-main}" + local scratch_dir + scratch_dir=$(mktemp -d /tmp/dispatcher-result-XXXXXX) + # shellcheck disable=SC2064 + trap "rm -rf '${scratch_dir}'" RETURN + + # Shallow clone of the ops repo — only the primary branch + if ! git clone --depth 1 --branch "$branch" \ + "$ops_clone_url" "$scratch_dir" 2>/dev/null; then + log "ERROR: Failed to clone ops repo for result commit (action ${action_id})" + return 1 + fi + + # Idempotency: skip if result.json already exists upstream + if [ -f "${scratch_dir}/${result_relpath}" ]; then + log "Result already exists upstream for ${action_id} — skipping commit" + return 0 + fi + + # Configure git identity as vault-bot + git -C "$scratch_dir" config user.name "vault-bot" + git -C "$scratch_dir" config user.email "vault-bot@disinto.local" # Truncate logs if too long (keep last 1000 chars) if [ ${#logs} -gt 1000 ]; then logs="${logs: -1000}" fi - # Write result JSON + # Write result JSON via jq (never string-interpolate into JSON) + mkdir -p "$(dirname "${scratch_dir}/${result_relpath}")" jq -n \ --arg id "$action_id" \ --argjson exit_code "$exit_code" \ --arg timestamp "$(date -u '+%Y-%m-%dT%H:%M:%SZ')" \ --arg logs "$logs" \ '{id: $id, exit_code: $exit_code, timestamp: $timestamp, logs: $logs}' \ - > "$result_file" + > "${scratch_dir}/${result_relpath}" - log "Result written: ${result_file}" + git -C "$scratch_dir" add "$result_relpath" + git -C "$scratch_dir" commit -q -m "vault: result for ${action_id}" + + # Push with retry on conflict (rebase-and-push pattern). + # Common case: admin merges another action PR between our clone and push. + local attempt + for attempt in 1 2 3; do + if git -C "$scratch_dir" push origin "$branch" 2>/dev/null; then + log "Result committed and pushed for ${action_id} (attempt ${attempt})" + return 0 + fi + + log "Push conflict for ${action_id} (attempt ${attempt}/3) — rebasing" + + if ! git -C "$scratch_dir" pull --rebase origin "$branch" 2>/dev/null; then + # Rebase conflict — check if result was pushed by another process + git -C "$scratch_dir" rebase --abort 2>/dev/null || true + if git -C "$scratch_dir" fetch origin "$branch" 2>/dev/null && \ + git -C "$scratch_dir" show "origin/${branch}:${result_relpath}" >/dev/null 2>&1; then + log "Result already exists upstream for ${action_id} (pushed by another process)" + return 0 + fi + fi + done + + log "ERROR: Failed to push result for ${action_id} after 3 attempts" + return 1 } -# Launch runner for the given action -# Usage: launch_runner -launch_runner() { - local toml_file="$1" - local action_id - action_id=$(basename "$toml_file" .toml) +# Write result file for an action via git push to the ops repo. +# Usage: write_result +write_result() { + local action_id="$1" + local exit_code="$2" + local logs="$3" - log "Launching runner for action: ${action_id}" + commit_result_via_git "$action_id" "$exit_code" "$logs" +} - # Validate TOML - if ! validate_action "$toml_file"; then - log "ERROR: Action validation failed for ${action_id}" - write_result "$action_id" 1 "Validation failed: see logs above" - return 1 - fi +# ----------------------------------------------------------------------------- +# Pluggable launcher backends +# ----------------------------------------------------------------------------- - # Check dispatch mode to determine if admin verification is needed - local dispatch_mode - dispatch_mode=$(get_dispatch_mode "$toml_file") +# _launch_runner_docker ACTION_ID SECRETS_CSV MOUNTS_CSV +# +# Builds and executes a `docker run` command for the vault runner. +# Secrets are resolved via load_secret (lib/env.sh). +# Returns: exit code of the docker run. Stdout/stderr are captured to a temp +# log file whose path is printed to stdout (caller reads it). +_launch_runner_docker() { + local action_id="$1" + local secrets_csv="$2" + local mounts_csv="$3" - if [ "$dispatch_mode" = "direct" ]; then - log "Action ${action_id}: tier=${VAULT_TIER:-unknown}, dispatch_mode=${dispatch_mode} — skipping admin merge verification (direct commit)" - else - # Verify admin merge for PR-based actions - log "Action ${action_id}: tier=${VAULT_TIER:-unknown}, dispatch_mode=${dispatch_mode} — verifying admin merge" - if ! verify_admin_merged "$toml_file"; then - log "ERROR: Admin merge verification failed for ${action_id}" - write_result "$action_id" 1 "Admin merge verification failed: see logs above" - return 1 - fi - log "Action ${action_id}: admin merge verified" - fi - - # Extract secrets from validated action - local secrets_array - secrets_array="${VAULT_ACTION_SECRETS:-}" - - # Build docker run command (self-contained, no compose context needed). - # The edge container has the Docker socket but not the host's compose project, - # so docker compose run would fail with exit 125. docker run is self-contained: - # the dispatcher knows the image, network, env vars, and entrypoint. local -a cmd=(docker run --rm --name "vault-runner-${action_id}" --network host @@ -451,29 +483,27 @@ launch_runner() { cmd+=(-v "${runtime_home}/.claude.json:/home/agent/.claude.json:ro") fi - # Add environment variables for secrets (if any declared) - if [ -n "$secrets_array" ]; then - for secret in $secrets_array; do + # Add environment variables for secrets (resolved via load_secret) + if [ -n "$secrets_csv" ]; then + local secret + for secret in $(echo "$secrets_csv" | tr ',' ' '); do secret=$(echo "$secret" | xargs) - if [ -n "$secret" ]; then - # Verify secret exists in vault - if [ -z "${!secret:-}" ]; then - log "ERROR: Secret '${secret}' not found in vault for action ${action_id}" - write_result "$action_id" 1 "Secret not found in vault: ${secret}" - return 1 - fi - cmd+=(-e "${secret}=${!secret}") + [ -n "$secret" ] || continue + local secret_val + secret_val=$(load_secret "$secret") || true + if [ -z "$secret_val" ]; then + log "ERROR: Secret '${secret}' could not be resolved for action ${action_id}" + write_result "$action_id" 1 "Secret not found: ${secret}" + return 1 fi + cmd+=(-e "${secret}=${secret_val}") done - else - log "Action ${action_id} has no secrets declared — runner will execute without extra env vars" fi - # Add volume mounts for file-based credentials (if any declared) - local mounts_array - mounts_array="${VAULT_ACTION_MOUNTS:-}" - if [ -n "$mounts_array" ]; then - for mount_alias in $mounts_array; do + # Add volume mounts for file-based credentials + if [ -n "$mounts_csv" ]; then + local mount_alias + for mount_alias in $(echo "$mounts_csv" | tr ',' ' '); do mount_alias=$(echo "$mount_alias" | xargs) [ -n "$mount_alias" ] || continue case "$mount_alias" in @@ -501,7 +531,7 @@ launch_runner() { # Image and entrypoint arguments: runner entrypoint + action-id cmd+=(disinto/agents:latest /home/agent/disinto/docker/runner/entrypoint-runner.sh "$action_id") - log "Running: docker run --rm vault-runner-${action_id} (secrets: ${secrets_array:-none}, mounts: ${mounts_array:-none})" + log "Running: docker run --rm vault-runner-${action_id} (secrets: ${secrets_csv:-none}, mounts: ${mounts_csv:-none})" # Create temp file for logs local log_file @@ -509,7 +539,6 @@ launch_runner() { trap 'rm -f "$log_file"' RETURN # Execute with array expansion (safe from shell injection) - # Capture stdout and stderr to log file "${cmd[@]}" > "$log_file" 2>&1 local exit_code=$? @@ -529,6 +558,137 @@ launch_runner() { return $exit_code } +# _launch_runner_nomad ACTION_ID SECRETS_CSV MOUNTS_CSV +# +# Nomad backend stub — will be implemented in migration Step 5. +_launch_runner_nomad() { + echo "nomad backend not yet implemented" >&2 + return 1 +} + +# Launch runner for the given action (backend-agnostic orchestrator) +# Usage: launch_runner +launch_runner() { + local toml_file="$1" + local action_id + action_id=$(basename "$toml_file" .toml) + + log "Launching runner for action: ${action_id}" + + # Validate TOML + if ! validate_action "$toml_file"; then + log "ERROR: Action validation failed for ${action_id}" + write_result "$action_id" 1 "Validation failed: see logs above" + return 1 + fi + + # Check dispatch mode to determine if admin verification is needed + local dispatch_mode + dispatch_mode=$(get_dispatch_mode "$toml_file") + + if [ "$dispatch_mode" = "direct" ]; then + log "Action ${action_id}: tier=${VAULT_TIER:-unknown}, dispatch_mode=${dispatch_mode} — skipping admin merge verification (direct commit)" + else + # Verify admin merge for PR-based actions + log "Action ${action_id}: tier=${VAULT_TIER:-unknown}, dispatch_mode=${dispatch_mode} — verifying admin merge" + if ! verify_admin_merged "$toml_file"; then + log "ERROR: Admin merge verification failed for ${action_id}" + write_result "$action_id" 1 "Admin merge verification failed: see logs above" + return 1 + fi + log "Action ${action_id}: admin merge verified" + fi + + # Build CSV lists from validated action metadata + local secrets_csv="" + if [ -n "${VAULT_ACTION_SECRETS:-}" ]; then + # Convert space-separated to comma-separated + secrets_csv=$(echo "${VAULT_ACTION_SECRETS}" | xargs | tr ' ' ',') + fi + + local mounts_csv="" + if [ -n "${VAULT_ACTION_MOUNTS:-}" ]; then + mounts_csv=$(echo "${VAULT_ACTION_MOUNTS}" | xargs | tr ' ' ',') + fi + + # Delegate to the selected backend + "_launch_runner_${DISPATCHER_BACKEND}" "$action_id" "$secrets_csv" "$mounts_csv" +} + +# ----------------------------------------------------------------------------- +# Pluggable sidecar launcher (reproduce / triage / verify) +# ----------------------------------------------------------------------------- + +# _dispatch_sidecar_docker CONTAINER_NAME ISSUE_NUM PROJECT_TOML IMAGE [FORMULA] +# +# Launches a sidecar container via docker run (background, pid-tracked). +# Prints the background PID to stdout. +_dispatch_sidecar_docker() { + local container_name="$1" + local issue_number="$2" + local project_toml="$3" + local image="$4" + local formula="${5:-}" + + local -a cmd=(docker run --rm + --name "${container_name}" + --network host + --security-opt apparmor=unconfined + -v /var/run/docker.sock:/var/run/docker.sock + -v agent-data:/home/agent/data + -v project-repos:/home/agent/repos + -e "FORGE_URL=${FORGE_URL}" + -e "FORGE_TOKEN=${FORGE_TOKEN}" + -e "FORGE_REPO=${FORGE_REPO}" + -e "PRIMARY_BRANCH=${PRIMARY_BRANCH:-main}" + -e DISINTO_CONTAINER=1 + ) + + # Set formula if provided + if [ -n "$formula" ]; then + cmd+=(-e "DISINTO_FORMULA=${formula}") + fi + + # Pass through ANTHROPIC_API_KEY if set + if [ -n "${ANTHROPIC_API_KEY:-}" ]; then + cmd+=(-e "ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}") + fi + + # Mount shared Claude config dir and ~/.ssh from the runtime user's home + local runtime_home="${HOME:-/home/debian}" + if [ -d "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}" ]; then + cmd+=(-v "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}") + cmd+=(-e "CLAUDE_CONFIG_DIR=${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config}") + fi + if [ -f "${runtime_home}/.claude.json" ]; then + cmd+=(-v "${runtime_home}/.claude.json:/home/agent/.claude.json:ro") + fi + if [ -d "${runtime_home}/.ssh" ]; then + cmd+=(-v "${runtime_home}/.ssh:/home/agent/.ssh:ro") + fi + if [ -f /usr/local/bin/claude ]; then + cmd+=(-v /usr/local/bin/claude:/usr/local/bin/claude:ro) + fi + + # Mount the project TOML into the container at a stable path + local container_toml="/home/agent/project.toml" + cmd+=(-v "${project_toml}:${container_toml}:ro") + + cmd+=("${image}" "$container_toml" "$issue_number") + + # Launch in background + "${cmd[@]}" & + echo $! +} + +# _dispatch_sidecar_nomad CONTAINER_NAME ISSUE_NUM PROJECT_TOML IMAGE [FORMULA] +# +# Nomad sidecar backend stub — will be implemented in migration Step 5. +_dispatch_sidecar_nomad() { + echo "nomad backend not yet implemented" >&2 + return 1 +} + # ----------------------------------------------------------------------------- # Reproduce dispatch — launch sidecar for bug-report issues # ----------------------------------------------------------------------------- @@ -607,52 +767,13 @@ dispatch_reproduce() { log "Dispatching reproduce-agent for issue #${issue_number} (project: ${project_toml})" - # Build docker run command using array (safe from injection) - local -a cmd=(docker run --rm - --name "disinto-reproduce-${issue_number}" - --network host - --security-opt apparmor=unconfined - -v /var/run/docker.sock:/var/run/docker.sock - -v agent-data:/home/agent/data - -v project-repos:/home/agent/repos - -e "FORGE_URL=${FORGE_URL}" - -e "FORGE_TOKEN=${FORGE_TOKEN}" - -e "FORGE_REPO=${FORGE_REPO}" - -e "PRIMARY_BRANCH=${PRIMARY_BRANCH:-main}" - -e DISINTO_CONTAINER=1 - ) + local bg_pid + bg_pid=$("_dispatch_sidecar_${DISPATCHER_BACKEND}" \ + "disinto-reproduce-${issue_number}" \ + "$issue_number" \ + "$project_toml" \ + "disinto-reproduce:latest") - # Pass through ANTHROPIC_API_KEY if set - if [ -n "${ANTHROPIC_API_KEY:-}" ]; then - cmd+=(-e "ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}") - fi - - # Mount shared Claude config dir and ~/.ssh from the runtime user's home if available - local runtime_home="${HOME:-/home/debian}" - if [ -d "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}" ]; then - cmd+=(-v "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}") - cmd+=(-e "CLAUDE_CONFIG_DIR=${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config}") - fi - if [ -f "${runtime_home}/.claude.json" ]; then - cmd+=(-v "${runtime_home}/.claude.json:/home/agent/.claude.json:ro") - fi - if [ -d "${runtime_home}/.ssh" ]; then - cmd+=(-v "${runtime_home}/.ssh:/home/agent/.ssh:ro") - fi - # Mount claude CLI binary if present on host - if [ -f /usr/local/bin/claude ]; then - cmd+=(-v /usr/local/bin/claude:/usr/local/bin/claude:ro) - fi - - # Mount the project TOML into the container at a stable path - local container_toml="/home/agent/project.toml" - cmd+=(-v "${project_toml}:${container_toml}:ro") - - cmd+=(disinto-reproduce:latest "$container_toml" "$issue_number") - - # Launch in background; write pid-file so we don't double-launch - "${cmd[@]}" & - local bg_pid=$! echo "$bg_pid" > "$(_reproduce_lockfile "$issue_number")" log "Reproduce container launched (pid ${bg_pid}) for issue #${issue_number}" } @@ -732,53 +853,14 @@ dispatch_triage() { log "Dispatching triage-agent for issue #${issue_number} (project: ${project_toml})" - # Build docker run command using array (safe from injection) - local -a cmd=(docker run --rm - --name "disinto-triage-${issue_number}" - --network host - --security-opt apparmor=unconfined - -v /var/run/docker.sock:/var/run/docker.sock - -v agent-data:/home/agent/data - -v project-repos:/home/agent/repos - -e "FORGE_URL=${FORGE_URL}" - -e "FORGE_TOKEN=${FORGE_TOKEN}" - -e "FORGE_REPO=${FORGE_REPO}" - -e "PRIMARY_BRANCH=${PRIMARY_BRANCH:-main}" - -e DISINTO_CONTAINER=1 - -e DISINTO_FORMULA=triage - ) + local bg_pid + bg_pid=$("_dispatch_sidecar_${DISPATCHER_BACKEND}" \ + "disinto-triage-${issue_number}" \ + "$issue_number" \ + "$project_toml" \ + "disinto-reproduce:latest" \ + "triage") - # Pass through ANTHROPIC_API_KEY if set - if [ -n "${ANTHROPIC_API_KEY:-}" ]; then - cmd+=(-e "ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}") - fi - - # Mount shared Claude config dir and ~/.ssh from the runtime user's home if available - local runtime_home="${HOME:-/home/debian}" - if [ -d "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}" ]; then - cmd+=(-v "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}") - cmd+=(-e "CLAUDE_CONFIG_DIR=${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config}") - fi - if [ -f "${runtime_home}/.claude.json" ]; then - cmd+=(-v "${runtime_home}/.claude.json:/home/agent/.claude.json:ro") - fi - if [ -d "${runtime_home}/.ssh" ]; then - cmd+=(-v "${runtime_home}/.ssh:/home/agent/.ssh:ro") - fi - # Mount claude CLI binary if present on host - if [ -f /usr/local/bin/claude ]; then - cmd+=(-v /usr/local/bin/claude:/usr/local/bin/claude:ro) - fi - - # Mount the project TOML into the container at a stable path - local container_toml="/home/agent/project.toml" - cmd+=(-v "${project_toml}:${container_toml}:ro") - - cmd+=(disinto-reproduce:latest "$container_toml" "$issue_number") - - # Launch in background; write pid-file so we don't double-launch - "${cmd[@]}" & - local bg_pid=$! echo "$bg_pid" > "$(_triage_lockfile "$issue_number")" log "Triage container launched (pid ${bg_pid}) for issue #${issue_number}" } @@ -934,53 +1016,14 @@ dispatch_verify() { log "Dispatching verification-agent for issue #${issue_number} (project: ${project_toml})" - # Build docker run command using array (safe from injection) - local -a cmd=(docker run --rm - --name "disinto-verify-${issue_number}" - --network host - --security-opt apparmor=unconfined - -v /var/run/docker.sock:/var/run/docker.sock - -v agent-data:/home/agent/data - -v project-repos:/home/agent/repos - -e "FORGE_URL=${FORGE_URL}" - -e "FORGE_TOKEN=${FORGE_TOKEN}" - -e "FORGE_REPO=${FORGE_REPO}" - -e "PRIMARY_BRANCH=${PRIMARY_BRANCH:-main}" - -e DISINTO_CONTAINER=1 - -e DISINTO_FORMULA=verify - ) + local bg_pid + bg_pid=$("_dispatch_sidecar_${DISPATCHER_BACKEND}" \ + "disinto-verify-${issue_number}" \ + "$issue_number" \ + "$project_toml" \ + "disinto-reproduce:latest" \ + "verify") - # Pass through ANTHROPIC_API_KEY if set - if [ -n "${ANTHROPIC_API_KEY:-}" ]; then - cmd+=(-e "ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}") - fi - - # Mount shared Claude config dir and ~/.ssh from the runtime user's home if available - local runtime_home="${HOME:-/home/debian}" - if [ -d "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}" ]; then - cmd+=(-v "${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}") - cmd+=(-e "CLAUDE_CONFIG_DIR=${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config}") - fi - if [ -f "${runtime_home}/.claude.json" ]; then - cmd+=(-v "${runtime_home}/.claude.json:/home/agent/.claude.json:ro") - fi - if [ -d "${runtime_home}/.ssh" ]; then - cmd+=(-v "${runtime_home}/.ssh:/home/agent/.ssh:ro") - fi - # Mount claude CLI binary if present on host - if [ -f /usr/local/bin/claude ]; then - cmd+=(-v /usr/local/bin/claude:/usr/local/bin/claude:ro) - fi - - # Mount the project TOML into the container at a stable path - local container_toml="/home/agent/project.toml" - cmd+=(-v "${project_toml}:${container_toml}:ro") - - cmd+=(disinto-reproduce:latest "$container_toml" "$issue_number") - - # Launch in background; write pid-file so we don't double-launch - "${cmd[@]}" & - local bg_pid=$! echo "$bg_pid" > "$(_verify_lockfile "$issue_number")" log "Verification container launched (pid ${bg_pid}) for issue #${issue_number}" } @@ -1002,10 +1045,25 @@ ensure_ops_repo() { # Main dispatcher loop main() { - log "Starting dispatcher..." + log "Starting dispatcher (backend=${DISPATCHER_BACKEND})..." log "Polling ops repo: ${VAULT_ACTIONS_DIR}" log "Admin users: ${ADMIN_USERS}" + # Validate backend selection at startup + case "$DISPATCHER_BACKEND" in + docker) ;; + nomad) + log "ERROR: nomad backend not yet implemented" + echo "nomad backend not yet implemented" >&2 + exit 1 + ;; + *) + log "ERROR: unknown DISPATCHER_BACKEND=${DISPATCHER_BACKEND}" + echo "unknown DISPATCHER_BACKEND=${DISPATCHER_BACKEND} (expected: docker, nomad)" >&2 + exit 1 + ;; + esac + while true; do # Refresh ops repo at the start of each poll cycle ensure_ops_repo diff --git a/docker/edge/entrypoint-edge.sh b/docker/edge/entrypoint-edge.sh index 7fc4f4f..1b5f94f 100755 --- a/docker/edge/entrypoint-edge.sh +++ b/docker/edge/entrypoint-edge.sh @@ -173,9 +173,40 @@ PROJECT_TOML="${PROJECT_TOML:-projects/disinto.toml}" sleep 1200 # 20 minutes done) & +# ── Load required secrets from secrets/*.enc (#777) ──────────────────── +# Edge container declares its required secrets; missing ones cause a hard fail. +_AGE_KEY_FILE="${HOME}/.config/sops/age/keys.txt" +_SECRETS_DIR="/opt/disinto/secrets" +EDGE_REQUIRED_SECRETS="CADDY_SSH_KEY CADDY_SSH_HOST CADDY_SSH_USER CADDY_ACCESS_LOG" + +_edge_decrypt_secret() { + local enc_path="${_SECRETS_DIR}/${1}.enc" + [ -f "$enc_path" ] || return 1 + age -d -i "$_AGE_KEY_FILE" "$enc_path" 2>/dev/null +} + +if [ -f "$_AGE_KEY_FILE" ] && [ -d "$_SECRETS_DIR" ]; then + _missing="" + for _secret_name in $EDGE_REQUIRED_SECRETS; do + _val=$(_edge_decrypt_secret "$_secret_name") || { _missing="${_missing} ${_secret_name}"; continue; } + export "$_secret_name=$_val" + done + if [ -n "$_missing" ]; then + echo "FATAL: required secrets missing from secrets/*.enc:${_missing}" >&2 + echo " Run 'disinto secrets add ' for each missing secret." >&2 + echo " If migrating from .env.vault.enc, run 'disinto secrets migrate-from-vault' first." >&2 + exit 1 + fi + echo "edge: loaded required secrets: ${EDGE_REQUIRED_SECRETS}" >&2 +else + echo "FATAL: age key (${_AGE_KEY_FILE}) or secrets dir (${_SECRETS_DIR}) not found — cannot load required secrets" >&2 + echo " Ensure age is installed and secrets/*.enc files are present." >&2 + exit 1 +fi + # Start daily engagement collection cron loop in background (#745) # Runs collect-engagement.sh daily at ~23:50 UTC via a sleep loop that -# calculates seconds until the next 23:50 window. SSH key from .env.vault.enc. +# calculates seconds until the next 23:50 window. SSH key from secrets/*.enc (#777). (while true; do # Calculate seconds until next 23:50 UTC _now=$(date -u +%s) @@ -186,26 +217,21 @@ done) & _sleep_secs=$(( _target - _now )) echo "edge: collect-engagement scheduled in ${_sleep_secs}s (next 23:50 UTC)" >&2 sleep "$_sleep_secs" - # Set CADDY_ACCESS_LOG so the script reads from the fetched local copy _fetch_log="/tmp/caddy-access-log-fetch.log" - if [ -n "${CADDY_SSH_KEY:-}" ]; then - _ssh_key_file=$(mktemp) - printf '%s\n' "$CADDY_SSH_KEY" > "$_ssh_key_file" - chmod 0600 "$_ssh_key_file" - scp -i "$_ssh_key_file" -o StrictHostKeyChecking=accept-new -o ConnectTimeout=10 -o BatchMode=yes \ - "${CADDY_SSH_USER:-debian}@${CADDY_SSH_HOST:-disinto.ai}:${CADDY_ACCESS_LOG:-/var/log/caddy/access.log}" \ - "$_fetch_log" 2>&1 | tee -a /opt/disinto-logs/collect-engagement.log || true - rm -f "$_ssh_key_file" - if [ -s "$_fetch_log" ]; then - CADDY_ACCESS_LOG="$_fetch_log" bash /opt/disinto/site/collect-engagement.sh 2>&1 \ - | tee -a /opt/disinto-logs/collect-engagement.log || true - else - echo "edge: collect-engagement: fetched log is empty, skipping parse" >&2 - fi - rm -f "$_fetch_log" + _ssh_key_file=$(mktemp) + printf '%s\n' "$CADDY_SSH_KEY" > "$_ssh_key_file" + chmod 0600 "$_ssh_key_file" + scp -i "$_ssh_key_file" -o StrictHostKeyChecking=accept-new -o ConnectTimeout=10 -o BatchMode=yes \ + "${CADDY_SSH_USER}@${CADDY_SSH_HOST}:${CADDY_ACCESS_LOG}" \ + "$_fetch_log" 2>&1 | tee -a /opt/disinto-logs/collect-engagement.log || true + rm -f "$_ssh_key_file" + if [ -s "$_fetch_log" ]; then + CADDY_ACCESS_LOG="$_fetch_log" bash /opt/disinto/site/collect-engagement.sh 2>&1 \ + | tee -a /opt/disinto-logs/collect-engagement.log || true else - echo "edge: collect-engagement: CADDY_SSH_KEY not set, skipping" >&2 + echo "edge: collect-engagement: fetched log is empty, skipping parse" >&2 fi + rm -f "$_fetch_log" done) & # Caddy as main process — run in foreground via wait so background jobs survive diff --git a/docker/reproduce/Dockerfile b/docker/reproduce/Dockerfile index 3192744..30bc75f 100644 --- a/docker/reproduce/Dockerfile +++ b/docker/reproduce/Dockerfile @@ -7,5 +7,8 @@ RUN apt-get update && apt-get install -y --no-install-recommends \ RUN useradd -m -u 1000 -s /bin/bash agent COPY docker/reproduce/entrypoint-reproduce.sh /entrypoint-reproduce.sh RUN chmod +x /entrypoint-reproduce.sh +VOLUME /home/agent/data +VOLUME /home/agent/repos + WORKDIR /home/agent ENTRYPOINT ["/entrypoint-reproduce.sh"] diff --git a/docs/VAULT.md b/docs/VAULT.md index 838c364..d927170 100644 --- a/docs/VAULT.md +++ b/docs/VAULT.md @@ -26,8 +26,8 @@ The `main` branch on the ops repo (`johba/disinto-ops`) is protected via Forgejo ## Vault PR Lifecycle -1. **Request** — Agent calls `lib/vault.sh:vault_request()` with action TOML content -2. **Validation** — TOML is validated against the schema in `vault/vault-env.sh` +1. **Request** — Agent calls `lib/action-vault.sh:vault_request()` with action TOML content +2. **Validation** — TOML is validated against the schema in `action-vault/vault-env.sh` 3. **PR Creation** — A PR is created on `disinto-ops` with: - Branch: `vault/` - Title: `vault: ` @@ -90,12 +90,12 @@ To verify the protection is working: - #73 — Vault redesign proposal - #74 — Vault action TOML schema -- #75 — Vault PR creation helper (`lib/vault.sh`) +- #75 — Vault PR creation helper (`lib/action-vault.sh`) - #76 — Dispatcher rewrite (poll for merged vault PRs) - #77 — Branch protection on ops repo (this issue) ## See Also -- [`lib/vault.sh`](../lib/vault.sh) — Vault PR creation helper -- [`vault/vault-env.sh`](../vault/vault-env.sh) — TOML validation +- [`lib/action-vault.sh`](../lib/action-vault.sh) — Vault PR creation helper +- [`action-vault/vault-env.sh`](../action-vault/vault-env.sh) — TOML validation - [`lib/branch-protection.sh`](../lib/branch-protection.sh) — Branch protection helper diff --git a/docs/agents-llama.md b/docs/agents-llama.md new file mode 100644 index 0000000..88622a7 --- /dev/null +++ b/docs/agents-llama.md @@ -0,0 +1,59 @@ +# agents-llama — Local-Qwen Agents + +The `agents-llama` service is an optional compose service that runs agents +backed by a local llama-server instance (e.g. Qwen) instead of the Anthropic +API. It uses the same Docker image as the main `agents` service but connects to +a local inference endpoint via `ANTHROPIC_BASE_URL`. + +Two profiles are available: + +| Profile | Service | Roles | Use case | +|---------|---------|-------|----------| +| _(default)_ | `agents-llama` | `dev` only | Conservative: single-role soak test | +| `agents-llama-all` | `agents-llama-all` | all 7 (review, dev, gardener, architect, planner, predictor, supervisor) | Pre-migration: validate every role on llama before Nomad cutover | + +## Enabling + +Set `ENABLE_LLAMA_AGENT=1` in `.env` (or `.env.enc`) and provide the required +credentials: + +```env +ENABLE_LLAMA_AGENT=1 +FORGE_TOKEN_LLAMA= +FORGE_PASS_LLAMA= +ANTHROPIC_BASE_URL=http://host.docker.internal:8081 # llama-server endpoint +``` + +Then regenerate the compose file (`disinto init ...`) and bring the stack up. + +### Running all 7 roles (agents-llama-all) + +```bash +docker compose --profile agents-llama-all up -d +``` + +This starts the `agents-llama-all` container with all 7 bot roles against the +local llama endpoint. The per-role forge tokens (`FORGE_REVIEW_TOKEN`, +`FORGE_GARDENER_TOKEN`, etc.) must be set in `.env` — they are the same tokens +used by the Claude-backed `agents` container. + +## Prerequisites + +- **llama-server** (or compatible OpenAI-API endpoint) running on the host, + reachable from inside Docker at the URL set in `ANTHROPIC_BASE_URL`. +- A Forgejo bot user (e.g. `dev-qwen`) with its own API token and password, + stored as `FORGE_TOKEN_LLAMA` / `FORGE_PASS_LLAMA`. + +## Behaviour + +- `agents-llama`: `AGENT_ROLES=dev` — only picks up dev work. +- `agents-llama-all`: `AGENT_ROLES=review,dev,gardener,architect,planner,predictor,supervisor` — runs all 7 roles. +- `CLAUDE_AUTOCOMPACT_PCT_OVERRIDE=60` — more aggressive compaction for smaller + context windows. +- Serialises on the llama-server's single KV cache (AD-002). + +## Disabling + +Set `ENABLE_LLAMA_AGENT=0` (or leave it unset) and regenerate. The service +block is omitted entirely from `docker-compose.yml`; the stack starts cleanly +without it. diff --git a/docs/mirror-bootstrap.md b/docs/mirror-bootstrap.md new file mode 100644 index 0000000..ca91d32 --- /dev/null +++ b/docs/mirror-bootstrap.md @@ -0,0 +1,59 @@ +# Mirror Bootstrap — Pull-Mirror Cutover Path + +How to populate an empty Forgejo repo from an external source using +`lib/mirrors.sh`'s `mirror_pull_register()`. + +## Prerequisites + +| Variable | Example | Purpose | +|---|---|---| +| `FORGE_URL` | `http://forgejo:3000` | Forgejo instance base URL | +| `FORGE_API_BASE` | `${FORGE_URL}/api/v1` | Global API base (set by `lib/env.sh`) | +| `FORGE_TOKEN` | (admin or org-owner token) | Must have `repo:create` scope | + +The target org/user must already exist on the Forgejo instance. + +## Command + +```bash +source lib/env.sh +source lib/mirrors.sh + +# Register a pull mirror — creates the repo and starts the first sync. +mirror_pull_register \ + "https://codeberg.org/johba/disinto.git" \ # source URL + "disinto-admin" \ # target owner + "disinto" \ # target repo name + "8h0m0s" # sync interval (optional, default 8h) +``` + +The function calls `POST /api/v1/repos/migrate` with `mirror: true`. +Forgejo creates the repo and immediately queues the first sync. + +## Verifying the sync + +```bash +# Check mirror status via API +forge_api GET "/repos/disinto-admin/disinto" | jq '.mirror, .mirror_interval' + +# Confirm content arrived — should list branches +forge_api GET "/repos/disinto-admin/disinto/branches" | jq '.[].name' +``` + +The first sync typically completes within a few seconds for small-to-medium +repos. For large repos, poll the branches endpoint until content appears. + +## Cutover scenario (Nomad migration) + +At cutover to the Nomad box: + +1. Stand up fresh Forgejo on the Nomad cluster (empty instance). +2. Create the `disinto-admin` org via `disinto init` or API. +3. Run `mirror_pull_register` pointing at the Codeberg source. +4. Wait for sync to complete (check branches endpoint). +5. Once content is confirmed, proceed with `disinto init` against the + now-populated repo — all subsequent `mirror_push` calls will push + to any additional mirrors configured in `projects/*.toml`. + +No manual `git clone` + `git push` step is needed. The Forgejo pull-mirror +handles the entire transfer. diff --git a/formulas/collect-engagement.toml b/formulas/collect-engagement.toml index fdfa65e..64ba54b 100644 --- a/formulas/collect-engagement.toml +++ b/formulas/collect-engagement.toml @@ -50,7 +50,7 @@ description = """ Fetch today's Caddy access log segment from the remote host using SCP. The SSH key is read from the environment (CADDY_SSH_KEY), which is -decrypted from .env.vault.enc by the dispatcher. It is NEVER hardcoded. +decrypted from secrets/CADDY_SSH_KEY.enc by the edge entrypoint. It is NEVER hardcoded. 1. Write the SSH key to a temporary file with restricted permissions: _ssh_key_file=$(mktemp) diff --git a/formulas/rent-a-human-caddy-ssh.toml b/formulas/rent-a-human-caddy-ssh.toml index 57dfc77..eb3aed1 100644 --- a/formulas/rent-a-human-caddy-ssh.toml +++ b/formulas/rent-a-human-caddy-ssh.toml @@ -79,28 +79,23 @@ AND set CADDY_ACCESS_LOG in the factory environment to match. [[steps]] id = "store-private-key" -title = "Add the private key to .env.vault.enc as CADDY_SSH_KEY" +title = "Add the private key as CADDY_SSH_KEY secret" needs = ["generate-keypair"] description = """ -Store the private key in the factory's encrypted vault secrets. +Store the private key in the factory's encrypted secrets store. -1. Read the private key: - cat caddy-collect +1. Add the private key using `disinto secrets add`: -2. Add it to .env.vault.enc (or .env.vault for plaintext fallback) as - CADDY_SSH_KEY. The key is multi-line, so use the base64-encoded form: + cat caddy-collect | disinto secrets add CADDY_SSH_KEY - echo "CADDY_SSH_KEY=$(base64 -w0 caddy-collect)" >> .env.vault.enc + This encrypts the key with age and stores it as secrets/CADDY_SSH_KEY.enc. - Or, if using SOPS-encrypted vault, decrypt first, add the variable, - then re-encrypt. - -3. IMPORTANT: After storing, securely delete the local private key file: +2. IMPORTANT: After storing, securely delete the local private key file: shred -u caddy-collect 2>/dev/null || rm -f caddy-collect rm -f caddy-collect.pub The public key is already installed on the Caddy host; the private key - now lives only in the vault. + now lives only in secrets/CADDY_SSH_KEY.enc. Never commit the private key to any git repository. """ @@ -109,20 +104,19 @@ Never commit the private key to any git repository. [[steps]] id = "store-caddy-host" -title = "Add the Caddy host address to .env.vault.enc as CADDY_HOST" +title = "Add the Caddy host details as secrets" needs = ["install-public-key"] description = """ -Store the Caddy host connection string so collect-engagement.sh knows +Store the Caddy connection details so collect-engagement.sh knows where to SSH. -1. Add to .env.vault.enc (or .env.vault for plaintext fallback): +1. Add each value using `disinto secrets add`: - echo "CADDY_HOST=user@caddy-host-ip-or-domain" >> .env.vault.enc + echo 'disinto.ai' | disinto secrets add CADDY_SSH_HOST + echo 'debian' | disinto secrets add CADDY_SSH_USER + echo '/var/log/caddy/access.log' | disinto secrets add CADDY_ACCESS_LOG - Replace user@caddy-host-ip-or-domain with the actual SSH user and host - (e.g. debian@203.0.113.42 or deploy@caddy.disinto.ai). - -2. If using SOPS, decrypt/add/re-encrypt as above. + Replace values with the actual SSH host, user, and log path for your setup. """ # ── Step 5: Test the connection ────────────────────────────────────────────── diff --git a/formulas/review-pr.toml b/formulas/review-pr.toml index fe62a89..ce6d2bf 100644 --- a/formulas/review-pr.toml +++ b/formulas/review-pr.toml @@ -213,7 +213,7 @@ should file a vault item instead of executing directly. **Exceptions** (do NOT flag these): - Code inside `vault/` — the vault system itself is allowed to handle secrets - References in comments or documentation explaining the architecture -- `bin/disinto` setup commands that manage `.env.vault.enc` and the `run` subcommand +- `bin/disinto` setup commands that manage `secrets/*.enc` and the `run` subcommand - Local operations (git push to forge, forge API calls with `FORGE_TOKEN`) ## 6. Re-review (if previous review is provided) diff --git a/formulas/run-gardener.toml b/formulas/run-gardener.toml index 7b0cdde..427aeb3 100644 --- a/formulas/run-gardener.toml +++ b/formulas/run-gardener.toml @@ -177,7 +177,7 @@ DUST (trivial — single-line edit, rename, comment, style, whitespace): VAULT (needs human decision or external resource): File a vault procurement item using vault_request(): - source "$(dirname "$0")/../lib/vault.sh" + source "$(dirname "$0")/../lib/action-vault.sh" TOML_CONTENT="# Vault action: context = \"\" unblocks = [\"#NNN\"] diff --git a/formulas/run-planner.toml b/formulas/run-planner.toml index ec6d6c8..aae72e8 100644 --- a/formulas/run-planner.toml +++ b/formulas/run-planner.toml @@ -243,7 +243,7 @@ needs = ["preflight"] [[steps]] id = "commit-ops-changes" -title = "Write tree, memory, and journal; commit and push" +title = "Write tree, memory, and journal; commit and push branch" description = """ ### 1. Write prerequisite tree Write to: $OPS_REPO_ROOT/prerequisites.md @@ -256,14 +256,16 @@ If (count - N) >= 5 or planner-memory.md missing, write to: Include: run counter marker, date, constraint focus, patterns, direction. Keep under 100 lines. Replace entire file. -### 3. Commit ops repo changes -Commit the ops repo changes (prerequisites, memory, vault items): +### 3. Commit ops repo changes to the planner branch +Commit the ops repo changes (prerequisites, memory, vault items) and push the +branch. Do NOT push directly to $PRIMARY_BRANCH — planner-run.sh will create a +PR and walk it to merge via review-bot. cd "$OPS_REPO_ROOT" git add prerequisites.md knowledge/planner-memory.md vault/pending/ git add -u if ! git diff --cached --quiet; then git commit -m "chore: planner run $(date -u +%Y-%m-%d)" - git push origin "$PRIMARY_BRANCH" + git push origin HEAD fi cd "$PROJECT_REPO_ROOT" diff --git a/formulas/run-predictor.toml b/formulas/run-predictor.toml index ddaa8a4..14364aa 100644 --- a/formulas/run-predictor.toml +++ b/formulas/run-predictor.toml @@ -125,8 +125,8 @@ For each weakness you identify, choose one: The prediction explains the theory. The vault PR triggers the proof after human approval. When the planner runs next, evidence is already there. - Vault dispatch (requires lib/vault.sh): - source "$PROJECT_REPO_ROOT/lib/vault.sh" + Vault dispatch (requires lib/action-vault.sh): + source "$PROJECT_REPO_ROOT/lib/action-vault.sh" TOML_CONTENT="id = \"predict--\" context = \"Test prediction #: — focus: \" @@ -154,7 +154,7 @@ tea is pre-configured with login "$TEA_LOGIN" and repo "$FORGE_REPO". --title "" --body "<body>" --labels "prediction/unreviewed" 2. Dispatch formula via vault (if exploiting): - source "$PROJECT_REPO_ROOT/lib/vault.sh" + source "$PROJECT_REPO_ROOT/lib/action-vault.sh" PR_NUM=$(vault_request "predict-NNN-<formula>" "$TOML_CONTENT") # See EXPLOIT section above for TOML_CONTENT format diff --git a/gardener/AGENTS.md b/gardener/AGENTS.md index 8d4c3af..1a2e08e 100644 --- a/gardener/AGENTS.md +++ b/gardener/AGENTS.md @@ -1,4 +1,4 @@ -<!-- last-reviewed: cdb0408466d875cdd893c712e3b3f0d0b909d4ef --> +<!-- last-reviewed: c363ee0aea2ae447daab28c2c850d6abefc8c6b5 --> # Gardener Agent **Role**: Backlog grooming — detect duplicate issues, missing acceptance @@ -32,7 +32,7 @@ the gardener runs as part of the polling loop alongside the planner, predictor, PR, reviewed alongside AGENTS.md changes, executed by gardener-run.sh after merge. **Environment variables consumed**: -- `FORGE_TOKEN`, `FORGE_GARDENER_TOKEN` (falls back to FORGE_TOKEN), `FORGE_REPO`, `FORGE_API`, `PROJECT_NAME`, `PROJECT_REPO_ROOT` +- `FORGE_TOKEN`, `FORGE_GARDENER_TOKEN` (falls back to FORGE_TOKEN), `FORGE_REPO`, `FORGE_API`, `PROJECT_NAME`, `PROJECT_REPO_ROOT`. `FORGE_TOKEN_OVERRIDE` is exported to `$FORGE_GARDENER_TOKEN` before sourcing env.sh so the gardener-bot identity survives re-sourcing (#762). - `PRIMARY_BRANCH`, `CLAUDE_MODEL` (set to sonnet by gardener-run.sh) **Lifecycle**: gardener-run.sh (invoked by polling loop every 6h, `check_active gardener`) → diff --git a/gardener/pending-actions.json b/gardener/pending-actions.json index fe51488..2c4c30f 100644 --- a/gardener/pending-actions.json +++ b/gardener/pending-actions.json @@ -1 +1,12 @@ -[] +[ + { + "action": "comment", + "issue": 623, + "body": "**Dependency check:** All blocking dependencies are now closed:\n- #620 ✓ closed\n- #621 ✓ closed \n- #622 ✓ closed\n\nPer the issue description: *\"Once #620/#621/#622 are green, this issue should fork into at least three backlog children: subpath routing + Forgejo ROOT_URL / Woodpecker HOST, disinto-chat container scaffold with OAuth gate, and Claude Code sandbox envelope + working-dir scoping.\"*\n\nThis vision issue is ready for the planner to decompose into backlog children." + }, + { + "action": "comment", + "issue": 758, + "body": "**Gardener flag:** This issue requires human admin action on Forgejo to resolve — changing branch protection settings on the ops repo. No automated formula can fix Forgejo admin settings.\n\nProposed options (from issue body):\n1. Add `planner-bot` to the merge whitelist in ops repo branch protection\n2. Remove branch protection from the ops repo (agents are primary writers)\n3. Create an admin-level service token for agents\n\nThis is blocking all ops repo writes (planner knowledge, sprint artifacts, vault items)." + } +] diff --git a/lib/AGENTS.md b/lib/AGENTS.md index 0f53ef8..86fd67a 100644 --- a/lib/AGENTS.md +++ b/lib/AGENTS.md @@ -1,4 +1,4 @@ -<!-- last-reviewed: cdb0408466d875cdd893c712e3b3f0d0b909d4ef --> +<!-- last-reviewed: c363ee0aea2ae447daab28c2c850d6abefc8c6b5 --> # Shared Helpers (`lib/`) All agents source `lib/env.sh` as their first action. Additional helpers are @@ -6,7 +6,7 @@ sourced as needed. | File | What it provides | Sourced by | |---|---|---| -| `lib/env.sh` | Loads `.env`, sets `FACTORY_ROOT`, exports project config (`FORGE_REPO`, `PROJECT_NAME`, etc.), defines `log()`, `forge_api()`, `forge_api_all()` (paginates all pages; accepts optional second TOKEN parameter, defaults to `$FORGE_TOKEN`; handles invalid/empty JSON responses gracefully — returns empty on parse error instead of crashing), `woodpecker_api()`, `wpdb()`, `memory_guard()` (skips agent if RAM < threshold). Auto-loads project TOML if `PROJECT_TOML` is set. Exports per-agent tokens (`FORGE_PLANNER_TOKEN`, `FORGE_GARDENER_TOKEN`, `FORGE_VAULT_TOKEN`, `FORGE_SUPERVISOR_TOKEN`, `FORGE_PREDICTOR_TOKEN`) — each falls back to `$FORGE_TOKEN` if not set. **Vault-only token guard (AD-006)**: `unset GITHUB_TOKEN CLAWHUB_TOKEN` so agents never hold external-action tokens — only the runner container receives them. **Container note**: when `DISINTO_CONTAINER=1`, `.env` is NOT re-sourced — compose already injects env vars (including `FORGE_URL=http://forgejo:3000`) and re-sourcing would clobber them. **Save/restore scope (#364)**: only `FORGE_URL` is preserved across `.env` re-sourcing (compose injects `http://forgejo:3000`, `.env` has `http://localhost:3000`). `FORGE_TOKEN` is NOT preserved so refreshed tokens in `.env` take effect immediately. **Required env var**: `FORGE_PASS` — bot password for git HTTP push (Forgejo 11.x rejects API tokens for `git push`, #361). **Hard preconditions (#674)**: `USER` and `HOME` must be exported by the entrypoint before sourcing. When `PROJECT_TOML` is set, `PROJECT_REPO_ROOT`, `PRIMARY_BRANCH`, and `OPS_REPO_ROOT` must also be set (by entrypoint or TOML). | Every agent | +| `lib/env.sh` | Loads `.env`, sets `FACTORY_ROOT`, exports project config (`FORGE_REPO`, `PROJECT_NAME`, etc.), defines `log()`, `forge_api()`, `forge_api_all()` (paginates all pages; accepts optional second TOKEN parameter, defaults to `$FORGE_TOKEN`; handles invalid/empty JSON responses gracefully — returns empty on parse error instead of crashing), `woodpecker_api()`, `wpdb()`, `memory_guard()` (skips agent if RAM < threshold), `load_secret()` (secret-source abstraction — see below). Auto-loads project TOML if `PROJECT_TOML` is set. Exports per-agent tokens (`FORGE_PLANNER_TOKEN`, `FORGE_GARDENER_TOKEN`, `FORGE_VAULT_TOKEN`, `FORGE_SUPERVISOR_TOKEN`, `FORGE_PREDICTOR_TOKEN`) — each falls back to `$FORGE_TOKEN` if not set. **Vault-only token guard (AD-006)**: `unset GITHUB_TOKEN CLAWHUB_TOKEN` so agents never hold external-action tokens — only the runner container receives them. **Container note**: when `DISINTO_CONTAINER=1`, `.env` is NOT re-sourced — compose already injects env vars (including `FORGE_URL=http://forgejo:3000`) and re-sourcing would clobber them. **Save/restore scope (#364)**: only `FORGE_URL` is preserved across `.env` re-sourcing (compose injects `http://forgejo:3000`, `.env` has `http://localhost:3000`). `FORGE_TOKEN` is NOT preserved so refreshed tokens in `.env` take effect immediately. **Per-agent token override (#762)**: agent run scripts export `FORGE_TOKEN_OVERRIDE=<agent-specific-token>` BEFORE sourcing `env.sh`; `env.sh` applies this override at lines 98-100, ensuring the correct identity survives any re-sourcing of `env.sh` by nested shells or `claude -p` invocations. **Required env var**: `FORGE_PASS` — bot password for git HTTP push (Forgejo 11.x rejects API tokens for `git push`, #361). **Hard preconditions (#674)**: `USER` and `HOME` must be exported by the entrypoint before sourcing. When `PROJECT_TOML` is set, `PROJECT_REPO_ROOT`, `PRIMARY_BRANCH`, and `OPS_REPO_ROOT` must also be set (by entrypoint or TOML). **`load_secret NAME [DEFAULT]` (#793)**: backend-agnostic secret resolution. Precedence: (1) `/secrets/<NAME>.env` — Nomad-rendered template, (2) current environment — already set by `.env.enc` / compose, (3) `secrets/<NAME>.enc` — age-encrypted per-key file (decrypted on demand, cached in process env), (4) DEFAULT or empty. Consumers call `$(load_secret GITHUB_TOKEN)` instead of `${GITHUB_TOKEN}` — identical behavior whether secrets come from Docker compose injection or Nomad Vault templates. | Every agent | | `lib/ci-helpers.sh` | `ci_passed()` — returns 0 if CI state is "success" (or no CI configured). `ci_required_for_pr()` — returns 0 if PR has code files (CI required), 1 if non-code only (CI not required). `is_infra_step()` — returns 0 if a single CI step failure matches infra heuristics (clone/git exit 128, any exit 137, log timeout patterns). `classify_pipeline_failure()` — returns "infra \<reason>" if any failed Woodpecker step matches infra heuristics via `is_infra_step()`, else "code". `ensure_priority_label()` — looks up (or creates) the `priority` label and returns its ID; caches in `_PRIORITY_LABEL_ID`. `ci_commit_status <sha>` — queries Woodpecker directly for CI state, falls back to forge commit status API. `ci_pipeline_number <sha>` — returns the Woodpecker pipeline number for a commit, falls back to parsing forge status `target_url`. `ci_promote <repo_id> <pipeline_num> <environment>` — promotes a pipeline to a named Woodpecker environment (vault-gated deployment: vault approves, vault-fire calls this — vault redesign in progress, see #73-#77). `ci_get_logs <pipeline_number> [--step <name>]` — reads CI logs from Woodpecker SQLite database via `lib/ci-log-reader.py`; outputs last 200 lines to stdout. Requires mounted woodpecker-data volume at /woodpecker-data. | dev-poll, review-poll, review-pr | | `lib/ci-debug.sh` | CLI tool for Woodpecker CI: `list`, `status`, `logs`, `failures` subcommands. Not sourced — run directly. | Humans / dev-agent (tool access) | | `lib/ci-log-reader.py` | Python tool: reads CI logs from Woodpecker SQLite database. `<pipeline_number> [--step <name>]` — returns last 200 lines from failed steps (or specified step). Used by `ci_get_logs()` in ci-helpers.sh. Requires `WOODPECKER_DATA_DIR` (default: /woodpecker-data). | ci-helpers.sh | @@ -14,7 +14,7 @@ sourced as needed. | `lib/parse-deps.sh` | Extracts dependency issue numbers from an issue body (stdin → stdout, one number per line). Matches `## Dependencies` / `## Depends on` / `## Blocked by` sections and inline `depends on #N` / `blocked by #N` patterns. Inline scan skips fenced code blocks to prevent false positives from code examples in issue bodies. Not sourced — executed via `bash lib/parse-deps.sh`. | dev-poll | | `lib/formula-session.sh` | `acquire_run_lock()`, `load_formula()`, `load_formula_or_profile()`, `build_context_block()`, `ensure_ops_repo()`, `ops_commit_and_push()`, `build_prompt_footer()`, `build_sdk_prompt_footer()`, `formula_worktree_setup()`, `formula_prepare_profile_context()`, `formula_lessons_block()`, `profile_write_journal()`, `profile_load_lessons()`, `ensure_profile_repo()`, `_profile_has_repo()`, `_count_undigested_journals()`, `_profile_digest_journals()`, `_profile_restore_lessons()`, `_profile_commit_and_push()`, `resolve_agent_identity()`, `build_graph_section()`, `build_scratch_instruction()`, `read_scratch_context()`, `cleanup_stale_crashed_worktrees()` — shared helpers for formula-driven polling-loop agents (lock, .profile repo management, prompt assembly, worktree setup). Memory guard is provided by `memory_guard()` in `lib/env.sh` (not duplicated here). `resolve_agent_identity()` — sets `FORGE_TOKEN`, `AGENT_IDENTITY`, `FORGE_REMOTE` from per-agent token env vars and FORGE_URL remote detection. `build_graph_section()` generates the structural-analysis section (runs `lib/build-graph.py`, formats JSON output) — previously duplicated in planner-run.sh and predictor-run.sh, now shared here. `cleanup_stale_crashed_worktrees()` — thin wrapper around `worktree_cleanup_stale()` from `lib/worktree.sh` (kept for backwards compatibility). **Journal digestion guards (#702)**: `_profile_digest_journals()` respects `PROFILE_DIGEST_TIMEOUT` (default 300s) and `PROFILE_DIGEST_MAX_BATCH` (default 5 journals per run); `_profile_restore_lessons()` restores the previous lessons-learned.md on digest failure. | planner-run.sh, predictor-run.sh, gardener-run.sh, supervisor-run.sh, dev-agent.sh | | `lib/guard.sh` | `check_active(agent_name)` — reads `$FACTORY_ROOT/state/.{agent_name}-active`; exits 0 (skip) if the file is absent. Factory is off by default — state files must be created to enable each agent. **Logs a message to stderr** when skipping (`[check_active] SKIP: state file not found`), so agent dropout is visible in loop logs. Sourced by dev-poll.sh, review-poll.sh, predictor-run.sh, supervisor-run.sh. | polling-loop entry points | -| `lib/mirrors.sh` | `mirror_push()` — pushes `$PRIMARY_BRANCH` + tags to all configured mirror remotes (fire-and-forget background pushes). Reads `MIRROR_NAMES` and `MIRROR_*` vars exported by `load-project.sh` from the `[mirrors]` TOML section. Failures are logged but never block the pipeline. Sourced by dev-poll.sh — called after every successful merge. | dev-poll.sh | +| `lib/mirrors.sh` | `mirror_push()` — pushes `$PRIMARY_BRANCH` + tags to all configured mirror remotes (fire-and-forget background pushes). Reads `MIRROR_NAMES` and `MIRROR_*` vars exported by `load-project.sh` from the `[mirrors]` TOML section. Failures are logged but never block the pipeline. `mirror_pull_register(clone_url, owner, repo_name, [interval])` — registers a Forgejo pull mirror via `POST /repos/migrate` with `mirror: true`. Creates the target repo and queues the first sync automatically. Works against empty Forgejo instances — no pre-existing content required. Used for Nomad migration cutover: point at Codeberg source, wait for sync, then proceed with `disinto init`. See [docs/mirror-bootstrap.md](../docs/mirror-bootstrap.md) for the full cutover path. Sourced by dev-poll.sh — called after every successful merge. | dev-poll.sh | | `lib/build-graph.py` | Python tool: parses VISION.md, prerequisites.md (from ops repo), AGENTS.md, formulas/*.toml, evidence/ (from ops repo), and forge issues/labels into a NetworkX DiGraph. Runs structural analyses (orphaned objectives, stale prerequisites, thin evidence, circular deps) and outputs a JSON report. Used by `review-pr.sh` (per-PR changed-file analysis) and `predictor-run.sh` (full-project analysis) to provide structural context to Claude. | review-pr.sh, predictor-run.sh | | `lib/secret-scan.sh` | `scan_for_secrets()` — detects potential secrets (API keys, bearer tokens, private keys, URLs with embedded credentials) in text; returns 1 if secrets found. `redact_secrets()` — replaces detected secret patterns with `[REDACTED]`. | issue-lifecycle.sh | | `lib/stack-lock.sh` | File-based lock protocol for singleton project stack access. `stack_lock_acquire(holder, project)` — polls until free, breaks stale heartbeats (>10 min old), claims lock. `stack_lock_release(project)` — deletes lock file. `stack_lock_check(project)` — inspect current lock state. `stack_lock_heartbeat(project)` — update heartbeat timestamp (callers must call every 2 min while holding). Lock files at `~/data/locks/<project>-stack.lock`. | docker/edge/dispatcher.sh, reproduce formula | @@ -22,7 +22,7 @@ sourced as needed. | `lib/worktree.sh` | Reusable git worktree management: `worktree_create(path, branch, [base_ref])` — create worktree, checkout base, fetch submodules. `worktree_recover(path, branch, [remote])` — detect existing worktree, reuse if on correct branch (sets `_WORKTREE_REUSED`), otherwise clean and recreate. `worktree_cleanup(path)` — `git worktree remove --force`, clear Claude Code project cache (`~/.claude/projects/` matching path). `worktree_cleanup_stale([max_age_hours])` — scan `/tmp` for orphaned worktrees older than threshold, skip preserved and active tmux worktrees, prune. `worktree_preserve(path, reason)` — mark worktree as preserved for debugging (writes `.worktree-preserved` marker, skipped by stale cleanup). | dev-agent.sh, supervisor-run.sh, planner-run.sh, predictor-run.sh, gardener-run.sh | | `lib/pr-lifecycle.sh` | Reusable PR lifecycle library: `pr_create()`, `pr_find_by_branch()`, `pr_poll_ci()`, `pr_poll_review()`, `pr_merge()`, `pr_is_merged()`, `pr_walk_to_merge()`, `build_phase_protocol_prompt()`. Requires `lib/ci-helpers.sh`. | dev-agent.sh (future) | | `lib/issue-lifecycle.sh` | Reusable issue lifecycle library: `issue_claim()` (add in-progress, remove backlog), `issue_release()` (remove in-progress, add backlog), `issue_block()` (post diagnostic comment with secret redaction, add blocked label), `issue_close()`, `issue_check_deps()` (parse deps, check transitive closure; sets `_ISSUE_BLOCKED_BY`, `_ISSUE_SUGGESTION`), `issue_suggest_next()` (find next unblocked backlog issue; sets `_ISSUE_NEXT`), `issue_post_refusal()` (structured refusal comment with dedup). Label IDs cached in globals on first lookup. Sources `lib/secret-scan.sh`. | dev-agent.sh (future) | -| `lib/vault.sh` | **Vault PR helper** — create vault action PRs on ops repo via Forgejo API (works from containers without SSH). `vault_request <action_id> <toml_content>` validates TOML (using `validate_vault_action` from `vault/vault-env.sh`), creates branch `vault/<action-id>`, writes `vault/actions/<action-id>.toml`, creates PR targeting `main` with title `vault: <action-id>` and body from context field, returns PR number. Idempotent: if PR exists, returns existing number. **Low-tier bypass**: if the action's `blast_radius` classifies as `low` (via `vault/classify.sh`), `vault_request` calls `_vault_commit_direct()` which commits directly to ops `main` using `FORGE_ADMIN_TOKEN` — no PR, no approval wait. Returns `0` (not a PR number) for direct commits. Requires `FORGE_TOKEN`, `FORGE_ADMIN_TOKEN` (low-tier only), `FORGE_URL`, `FORGE_REPO`, `FORGE_OPS_REPO`. Uses the calling agent's own token (saves/restores `FORGE_TOKEN` around sourcing `vault-env.sh`), so approval workflow respects individual agent identities. | dev-agent (vault actions), future vault dispatcher | +| `lib/action-vault.sh` | **Vault PR helper** — create vault action PRs on ops repo via Forgejo API (works from containers without SSH). `vault_request <action_id> <toml_content>` validates TOML (using `validate_vault_action` from `action-vault/vault-env.sh`), creates branch `vault/<action-id>`, writes `vault/actions/<action-id>.toml`, creates PR targeting `main` with title `vault: <action-id>` and body from context field, returns PR number. Idempotent: if PR exists, returns existing number. **Low-tier bypass**: if the action's `blast_radius` classifies as `low` (via `action-vault/classify.sh`), `vault_request` calls `_vault_commit_direct()` which commits directly to ops `main` using `FORGE_ADMIN_TOKEN` — no PR, no approval wait. Returns `0` (not a PR number) for direct commits. Requires `FORGE_TOKEN`, `FORGE_ADMIN_TOKEN` (low-tier only), `FORGE_URL`, `FORGE_REPO`, `FORGE_OPS_REPO`. Uses the calling agent's own token (saves/restores `FORGE_TOKEN` around sourcing `vault-env.sh`), so approval workflow respects individual agent identities. | dev-agent (vault actions), future vault dispatcher | | `lib/branch-protection.sh` | Branch protection helpers for Forgejo repos. `setup_vault_branch_protection()` — configures admin-only merge protection on main (require 1 approval, restrict merge to admin role, block direct pushes). `setup_profile_branch_protection()` — same protection for `.profile` repos. `verify_branch_protection()` — checks protection is correctly configured. `remove_branch_protection()` — removes protection (cleanup/testing). Handles race condition after initial push: retries with backoff if Forgejo hasn't processed the branch yet. Requires `FORGE_TOKEN`, `FORGE_URL`, `FORGE_OPS_REPO`. | bin/disinto (hire-an-agent) | | `lib/agent-sdk.sh` | `agent_run([--resume SESSION_ID] [--worktree DIR] PROMPT)` — one-shot `claude -p` invocation with session persistence. Saves session ID to `SID_FILE`, reads it back on resume. `agent_recover_session()` — restore previous session ID from `SID_FILE` on startup. **Nudge guard**: skips nudge injection if the worktree is clean and no push is expected, preventing spurious re-invocations. Callers must define `SID_FILE`, `LOGFILE`, and `log()` before sourcing. **Concurrency**: external `flock` on `session.lock` is gated behind `CLAUDE_EXTERNAL_LOCK=1` (default off). When unset, each container's per-session `CLAUDE_CONFIG_DIR` isolation lets Claude Code's native lockfile handle OAuth refresh — no external serialization needed. Set `CLAUDE_EXTERNAL_LOCK=1` to re-enable the old flock wrapper as a rollback mechanism. See [`docs/CLAUDE-AUTH-CONCURRENCY.md`](../docs/CLAUDE-AUTH-CONCURRENCY.md) and AD-002 (#647). | formula-driven agents (dev-agent, planner-run, predictor-run, gardener-run) | | `lib/forge-setup.sh` | `setup_forge()` — Forgejo instance provisioning: creates admin user, bot accounts, org, repos (code + ops), configures webhooks, sets repo topics. Extracted from `bin/disinto`. Requires `FORGE_URL`, `FORGE_TOKEN`, `FACTORY_ROOT`. **Password storage (#361)**: after creating each bot account, stores its password in `.env` as `FORGE_<BOT>_PASS` (e.g. `FORGE_PASS`, `FORGE_REVIEW_PASS`, etc.) for use by `forge-push.sh`. | bin/disinto (init) | @@ -30,6 +30,8 @@ sourced as needed. | `lib/git-creds.sh` | Shared git credential helper configuration. `configure_git_creds([HOME_DIR] [RUN_AS_CMD])` — writes a static credential helper script and configures git globally to use password-based HTTP auth (Forgejo 11.x rejects API tokens for `git push`, #361). **Retry on cold boot (#741)**: resolves bot username from `FORGE_TOKEN` with 5 retries (exponential backoff 1-5s); fails loudly and returns 1 if Forgejo is unreachable — never falls back to a wrong hardcoded default (exports `BOT_USER` on success). `repair_baked_cred_urls([--as RUN_AS_CMD] DIR ...)` — rewrites any git remote URLs that have credentials baked in to use clean URLs instead; uses `safe.directory` bypass for root-owned repos (#671). Requires `FORGE_PASS`, `FORGE_URL`, `FORGE_TOKEN`. | entrypoints (agents, edge) | | `lib/ops-setup.sh` | `setup_ops_repo()` — creates ops repo on Forgejo if it doesn't exist, configures bot collaborators, clones/initializes ops repo locally, seeds directory structure (vault, knowledge, evidence, sprints). Evidence subdirectories seeded: engagement/, red-team/, holdout/, evolution/, user-test/. Also seeds sprints/ for architect output. Exports `_ACTUAL_OPS_SLUG`. `migrate_ops_repo(ops_root, [primary_branch])` — idempotent migration helper that seeds missing directories and .gitkeep files on existing ops repos (pre-#407 deployments). | bin/disinto (init) | | `lib/ci-setup.sh` | `_install_cron_impl()` — installs crontab entries for bare-metal deployments (compose mode uses polling loop instead). `_create_forgejo_oauth_app()` — generic helper to create an OAuth2 app on Forgejo (shared by Woodpecker and chat). `_create_woodpecker_oauth_impl()` — creates Woodpecker OAuth2 app (thin wrapper). `_create_chat_oauth_impl()` — creates disinto-chat OAuth2 app, writes `CHAT_OAUTH_CLIENT_ID`/`CHAT_OAUTH_CLIENT_SECRET` to `.env` (#708). `_generate_woodpecker_token_impl()` — auto-generates WOODPECKER_TOKEN via OAuth2 flow. `_activate_woodpecker_repo_impl()` — activates repo in Woodpecker. All gated by `_load_ci_context()` which validates required env vars. | bin/disinto (init) | -| `lib/generators.sh` | Template generation for `disinto init`: `generate_compose()` — docker-compose.yml (uses `codeberg.org/forgejo/forgejo:11.0` tag; adds `security_opt: [apparmor:unconfined]` to all services for rootless container compatibility; Forgejo includes a healthcheck so dependent services use `condition: service_healthy` — fixes cold-start races, #665; adds `chat` service block with isolated `chat-config` named volume and `CHAT_HISTORY_DIR` bind-mount for per-user NDJSON history persistence (#710); injects `FORWARD_AUTH_SECRET` for Caddy↔chat defense-in-depth auth (#709); cost-cap env vars `CHAT_MAX_REQUESTS_PER_HOUR`, `CHAT_MAX_REQUESTS_PER_DAY`, `CHAT_MAX_TOKENS_PER_DAY` (#711); subdomain fallback comment for `EDGE_TUNNEL_FQDN_*` vars (#713); all `depends_on` now use `condition: service_healthy/started` instead of bare service names), `generate_caddyfile()` — Caddyfile (routes: `/forge/*` → forgejo:3000, `/woodpecker/*` → woodpecker:8000, `/staging/*` → staging:80; `/chat/login` and `/chat/oauth/callback` bypass `forward_auth` so unauthenticated users can reach the OAuth flow; `/chat/*` gated by `forward_auth` on `chat:8080/chat/auth/verify` which stamps `X-Forwarded-User` (#709); root `/` redirects to `/forge/`), `generate_staging_index()` — staging index, `generate_deploy_pipelines()` — Woodpecker deployment pipeline configs. Requires `FACTORY_ROOT`, `PROJECT_NAME`, `PRIMARY_BRANCH`. | bin/disinto (init) | +| `lib/generators.sh` | Template generation for `disinto init`: `generate_compose()` — docker-compose.yml (uses `codeberg.org/forgejo/forgejo:11.0` tag; adds `security_opt: [apparmor:unconfined]` to all services for rootless container compatibility; Forgejo includes a healthcheck so dependent services use `condition: service_healthy` — fixes cold-start races, #665; adds `chat` service block with isolated `chat-config` named volume and `CHAT_HISTORY_DIR` bind-mount for per-user NDJSON history persistence (#710); injects `FORWARD_AUTH_SECRET` for Caddy↔chat defense-in-depth auth (#709); cost-cap env vars `CHAT_MAX_REQUESTS_PER_HOUR`, `CHAT_MAX_REQUESTS_PER_DAY`, `CHAT_MAX_TOKENS_PER_DAY` (#711); subdomain fallback comment for `EDGE_TUNNEL_FQDN_*` vars (#713); all `depends_on` now use `condition: service_healthy/started` instead of bare service names; all services now include `restart: unless-stopped` including the edge service — #768; agents service now uses `image: ghcr.io/disinto/agents:${DISINTO_IMAGE_TAG:-latest}` instead of `build:` (#429); `WOODPECKER_PLUGINS_PRIVILEGED` env var added to woodpecker service (#779); agents-llama conditional block gated on `ENABLE_LLAMA_AGENT=1` (#769); `agents-llama-all` compose service (profile `agents-llama-all`, all 7 roles: review,dev,gardener,architect,planner,predictor,supervisor) added by #801; agents service gains volume mounts for `./projects`, `./.env`, `./state`), `generate_caddyfile()` — Caddyfile (routes: `/forge/*` → forgejo:3000, `/woodpecker/*` → woodpecker:8000, `/staging/*` → staging:80; `/chat/login` and `/chat/oauth/callback` bypass `forward_auth` so unauthenticated users can reach the OAuth flow; `/chat/*` gated by `forward_auth` on `chat:8080/chat/auth/verify` which stamps `X-Forwarded-User` (#709); root `/` redirects to `/forge/`), `generate_staging_index()` — staging index, `generate_deploy_pipelines()` — Woodpecker deployment pipeline configs. Requires `FACTORY_ROOT`, `PROJECT_NAME`, `PRIMARY_BRANCH`. | bin/disinto (init) | +| `lib/sprint-filer.sh` | Post-merge sub-issue filer for sprint PRs. Invoked by the `.woodpecker/ops-filer.yml` pipeline after a sprint PR merges to ops repo `main`. Parses `<!-- filer:begin --> ... <!-- filer:end -->` blocks from sprint PR bodies to extract sub-issue definitions, creates them on the project repo using `FORGE_FILER_TOKEN` (narrow-scope `filer-bot` identity with `issues:write` only), adds `in-progress` label to the parent vision issue, and handles vision lifecycle closure when all sub-issues are closed. Uses `filer_api_all()` for paginated fetches. Idempotent: uses `<!-- decomposed-from: #<vision>, sprint: <slug>, id: <id> -->` markers to skip already-filed issues. Requires `FORGE_FILER_TOKEN`, `FORGE_API`, `FORGE_API_BASE`, `FORGE_OPS_REPO`. | `.woodpecker/ops-filer.yml` (CI pipeline on ops repo) | | `lib/hire-agent.sh` | `disinto_hire_an_agent()` — user creation, `.profile` repo setup, formula copying, branch protection, and state marker creation for hiring a new agent. Requires `FORGE_URL`, `FORGE_TOKEN`, `FACTORY_ROOT`, `PROJECT_NAME`. Extracted from `bin/disinto`. | bin/disinto (hire) | | `lib/release.sh` | `disinto_release()` — vault TOML creation, branch setup on ops repo, PR creation, and auto-merge request for a versioned release. `_assert_release_globals()` validates required env vars. Requires `FORGE_URL`, `FORGE_TOKEN`, `FORGE_OPS_REPO`, `FACTORY_ROOT`, `PRIMARY_BRANCH`. Extracted from `bin/disinto`. | bin/disinto (release) | +| `lib/hvault.sh` | HashiCorp Vault helper module. `hvault_kv_get(PATH, [KEY])` — read KV v2 secret, optionally extract one key. `hvault_kv_put(PATH, KEY=VAL ...)` — write KV v2 secret. `hvault_kv_list(PATH)` — list keys at a KV path. `hvault_policy_apply(NAME, FILE)` — idempotent policy upsert. `hvault_jwt_login(ROLE, JWT)` — exchange JWT for short-lived token. `hvault_token_lookup()` — returns TTL/policies/accessor for current token. All functions use `VAULT_ADDR` + `VAULT_TOKEN` from env (fallback: `/etc/vault.d/root.token`), emit structured JSON errors to stderr on failure. Tests: `tests/lib-hvault.bats` (requires `vault server -dev`). | Not sourced at runtime yet — pure scaffolding for Nomad+Vault migration (#799) | diff --git a/lib/vault.sh b/lib/action-vault.sh similarity index 97% rename from lib/vault.sh rename to lib/action-vault.sh index 484fd57..6348cc6 100644 --- a/lib/vault.sh +++ b/lib/action-vault.sh @@ -1,9 +1,9 @@ #!/usr/bin/env bash -# vault.sh — Helper for agents to create vault PRs on ops repo +# action-vault.sh — Helper for agents to create vault PRs on ops repo # # Source after lib/env.sh: # source "$(dirname "$0")/../lib/env.sh" -# source "$(dirname "$0")/lib/vault.sh" +# source "$(dirname "$0")/lib/action-vault.sh" # # Required globals: FORGE_TOKEN, FORGE_URL, FORGE_REPO, FORGE_OPS_REPO # Optional: OPS_REPO_ROOT (local path for ops repo) @@ -12,7 +12,7 @@ # vault_request <action_id> <toml_content> — Create vault PR, return PR number # # The function: -# 1. Validates TOML content using validate_vault_action() from vault/vault-env.sh +# 1. Validates TOML content using validate_vault_action() from action-vault/vault-env.sh # 2. Creates a branch on the ops repo: vault/<action-id> # 3. Writes TOML to vault/actions/<action-id>.toml on that branch # 4. Creates PR targeting main with title "vault: <action-id>" @@ -133,7 +133,7 @@ vault_request() { printf '%s' "$toml_content" > "$tmp_toml" # Source vault-env.sh for validate_vault_action - local vault_env="${FACTORY_ROOT:-$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)}/vault/vault-env.sh" + local vault_env="${FACTORY_ROOT:-$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)}/action-vault/vault-env.sh" if [ ! -f "$vault_env" ]; then echo "ERROR: vault-env.sh not found at $vault_env" >&2 return 1 @@ -161,7 +161,7 @@ vault_request() { ops_api="$(_vault_ops_api)" # Classify the action to determine if PR bypass is allowed - local classify_script="${FACTORY_ROOT:-$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)}/vault/classify.sh" + local classify_script="${FACTORY_ROOT:-$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)}/action-vault/classify.sh" local vault_tier vault_tier=$("$classify_script" "${VAULT_ACTION_FORMULA:-}" "${VAULT_BLAST_RADIUS_OVERRIDE:-}") || { # Classification failed, default to high tier (require PR) diff --git a/lib/env.sh b/lib/env.sh index e91412c..85acb34 100755 --- a/lib/env.sh +++ b/lib/env.sh @@ -158,8 +158,8 @@ export WOODPECKER_SERVER="${WOODPECKER_SERVER:-http://localhost:8000}" export CLAUDE_TIMEOUT="${CLAUDE_TIMEOUT:-7200}" # Vault-only token guard (#745): external-action tokens (GITHUB_TOKEN, CLAWHUB_TOKEN) -# must NEVER be available to agents. They live in .env.vault.enc and are injected -# only into the ephemeral runner container at fire time. Unset them here so +# must NEVER be available to agents. They live in secrets/*.enc and are decrypted +# only into the ephemeral runner container at fire time (#777). Unset them here so # even an accidental .env inclusion cannot leak them into agent sessions. unset GITHUB_TOKEN 2>/dev/null || true unset CLAWHUB_TOKEN 2>/dev/null || true @@ -313,6 +313,68 @@ memory_guard() { fi } +# ============================================================================= +# SECRET LOADING ABSTRACTION +# ============================================================================= +# load_secret NAME [DEFAULT] +# +# Resolves a secret value using the following precedence: +# 1. /secrets/<NAME>.env — Nomad-rendered template (future) +# 2. Current environment — already set by .env.enc, compose, etc. +# 3. secrets/<NAME>.enc — age-encrypted per-key file (decrypted on demand) +# 4. DEFAULT (or empty) +# +# Prints the resolved value to stdout. Caches age-decrypted values in the +# process environment so subsequent calls are free. +# ============================================================================= +load_secret() { + local name="$1" + local default="${2:-}" + + # 1. Nomad-rendered template (future: Nomad writes /secrets/<NAME>.env) + local nomad_path="/secrets/${name}.env" + if [ -f "$nomad_path" ]; then + # Source into a subshell to extract just the value + local _nomad_val + _nomad_val=$( + set -a + # shellcheck source=/dev/null + source "$nomad_path" + set +a + printf '%s' "${!name:-}" + ) + if [ -n "$_nomad_val" ]; then + export "$name=$_nomad_val" + printf '%s' "$_nomad_val" + return 0 + fi + fi + + # 2. Already in environment (set by .env.enc, compose injection, etc.) + if [ -n "${!name:-}" ]; then + printf '%s' "${!name}" + return 0 + fi + + # 3. Age-encrypted per-key file: secrets/<NAME>.enc (#777) + local _age_key="${HOME}/.config/sops/age/keys.txt" + local _enc_path="${FACTORY_ROOT}/secrets/${name}.enc" + if [ -f "$_enc_path" ] && [ -f "$_age_key" ] && command -v age &>/dev/null; then + local _dec_val + if _dec_val=$(age -d -i "$_age_key" "$_enc_path" 2>/dev/null) && [ -n "$_dec_val" ]; then + export "$name=$_dec_val" + printf '%s' "$_dec_val" + return 0 + fi + fi + + # 4. Default (or empty) + if [ -n "$default" ]; then + printf '%s' "$default" + fi + return 0 +} + # Source tea helpers (available when tea binary is installed) if command -v tea &>/dev/null; then # shellcheck source=tea-helpers.sh diff --git a/lib/forge-setup.sh b/lib/forge-setup.sh index b925103..2b7b697 100644 --- a/lib/forge-setup.sh +++ b/lib/forge-setup.sh @@ -31,8 +31,9 @@ _load_init_context() { # Execute a command in the Forgejo container (for admin operations) _forgejo_exec() { local use_bare="${DISINTO_BARE:-false}" + local cname="${FORGEJO_CONTAINER_NAME:-disinto-forgejo}" if [ "$use_bare" = true ]; then - docker exec -u git disinto-forgejo "$@" + docker exec -u git "$cname" "$@" else docker compose -f "${FACTORY_ROOT}/docker-compose.yml" exec -T -u git forgejo "$@" fi @@ -94,11 +95,12 @@ setup_forge() { # Bare-metal mode: standalone docker run mkdir -p "${FORGEJO_DATA_DIR}" - if docker ps -a --format '{{.Names}}' | grep -q '^disinto-forgejo$'; then - docker start disinto-forgejo >/dev/null 2>&1 || true + local cname="${FORGEJO_CONTAINER_NAME:-disinto-forgejo}" + if docker ps -a --format '{{.Names}}' | grep -q "^${cname}$"; then + docker start "$cname" >/dev/null 2>&1 || true else docker run -d \ - --name disinto-forgejo \ + --name "$cname" \ --restart unless-stopped \ -p "${forge_port}:3000" \ -p 2222:22 \ @@ -210,8 +212,8 @@ setup_forge() { # Create human user (disinto-admin) as site admin if it doesn't exist local human_user="disinto-admin" - local human_pass - human_pass="admin-$(head -c 16 /dev/urandom | base64 | tr -dc 'a-zA-Z0-9' | head -c 20)" + # human_user == admin_user; reuse admin_pass for basic-auth operations + local human_pass="$admin_pass" if ! curl -sf --max-time 5 -H "Authorization: token ${FORGE_TOKEN:-}" "${forge_url}/api/v1/users/${human_user}" >/dev/null 2>&1; then echo "Creating human user: ${human_user}" @@ -243,63 +245,89 @@ setup_forge() { echo "Human user: ${human_user} (already exists)" fi - # Delete existing admin token if present (token sha1 is only returned at creation time) - local existing_token_id - existing_token_id=$(curl -sf \ - -u "${admin_user}:${admin_pass}" \ - "${forge_url}/api/v1/users/${admin_user}/tokens" 2>/dev/null \ - | jq -r '.[] | select(.name == "disinto-admin-token") | .id') || existing_token_id="" - if [ -n "$existing_token_id" ]; then - curl -sf -X DELETE \ - -u "${admin_user}:${admin_pass}" \ - "${forge_url}/api/v1/users/${admin_user}/tokens/${existing_token_id}" >/dev/null 2>&1 || true + # Preserve admin token if already stored in .env (idempotent re-run) + local admin_token="" + if _token_exists_in_env "FORGE_ADMIN_TOKEN" "$env_file" && [ "$rotate_tokens" = false ]; then + admin_token=$(grep '^FORGE_ADMIN_TOKEN=' "$env_file" | head -1 | cut -d= -f2-) + [ -n "$admin_token" ] && echo "Admin token: preserved (use --rotate-tokens to force)" fi - # Create admin token (fresh, so sha1 is returned) - local admin_token - admin_token=$(curl -sf -X POST \ - -u "${admin_user}:${admin_pass}" \ - -H "Content-Type: application/json" \ - "${forge_url}/api/v1/users/${admin_user}/tokens" \ - -d '{"name":"disinto-admin-token","scopes":["all"]}' 2>/dev/null \ - | jq -r '.sha1 // empty') || admin_token="" - if [ -z "$admin_token" ]; then - echo "Error: failed to obtain admin API token" >&2 - exit 1 - fi - - # Get or create human user token - local human_token="" - # Delete existing human token if present (token sha1 is only returned at creation time) - local existing_human_token_id - existing_human_token_id=$(curl -sf \ - -u "${human_user}:${human_pass}" \ - "${forge_url}/api/v1/users/${human_user}/tokens" 2>/dev/null \ - | jq -r '.[] | select(.name == "disinto-human-token") | .id') || existing_human_token_id="" - if [ -n "$existing_human_token_id" ]; then - curl -sf -X DELETE \ - -u "${human_user}:${human_pass}" \ - "${forge_url}/api/v1/users/${human_user}/tokens/${existing_human_token_id}" >/dev/null 2>&1 || true - fi - - # Create human token (fresh, so sha1 is returned) - human_token=$(curl -sf -X POST \ - -u "${human_user}:${human_pass}" \ - -H "Content-Type: application/json" \ - "${forge_url}/api/v1/users/${human_user}/tokens" \ - -d '{"name":"disinto-human-token","scopes":["all"]}' 2>/dev/null \ - | jq -r '.sha1 // empty') || human_token="" - - if [ -n "$human_token" ]; then - # Store human token in .env - if grep -q '^HUMAN_TOKEN=' "$env_file" 2>/dev/null; then - sed -i "s|^HUMAN_TOKEN=.*|HUMAN_TOKEN=${human_token}|" "$env_file" - else - printf 'HUMAN_TOKEN=%s\n' "$human_token" >> "$env_file" + # Delete existing admin token if present (token sha1 is only returned at creation time) + local existing_token_id + existing_token_id=$(curl -sf \ + -u "${admin_user}:${admin_pass}" \ + "${forge_url}/api/v1/users/${admin_user}/tokens" 2>/dev/null \ + | jq -r '.[] | select(.name == "disinto-admin-token") | .id') || existing_token_id="" + if [ -n "$existing_token_id" ]; then + curl -sf -X DELETE \ + -u "${admin_user}:${admin_pass}" \ + "${forge_url}/api/v1/users/${admin_user}/tokens/${existing_token_id}" >/dev/null 2>&1 || true + fi + + # Create admin token (fresh, so sha1 is returned) + admin_token=$(curl -sf -X POST \ + -u "${admin_user}:${admin_pass}" \ + -H "Content-Type: application/json" \ + "${forge_url}/api/v1/users/${admin_user}/tokens" \ + -d '{"name":"disinto-admin-token","scopes":["all"]}' 2>/dev/null \ + | jq -r '.sha1 // empty') || admin_token="" + + if [ -z "$admin_token" ]; then + echo "Error: failed to obtain admin API token" >&2 + exit 1 + fi + + # Store admin token for idempotent re-runs + if grep -q '^FORGE_ADMIN_TOKEN=' "$env_file" 2>/dev/null; then + sed -i "s|^FORGE_ADMIN_TOKEN=.*|FORGE_ADMIN_TOKEN=${admin_token}|" "$env_file" + else + printf 'FORGE_ADMIN_TOKEN=%s\n' "$admin_token" >> "$env_file" + fi + echo "Admin token: generated and saved (FORGE_ADMIN_TOKEN)" + fi + + # Get or create human user token (human_user == admin_user; use admin_pass) + local human_token="" + if _token_exists_in_env "HUMAN_TOKEN" "$env_file" && [ "$rotate_tokens" = false ]; then + human_token=$(grep '^HUMAN_TOKEN=' "$env_file" | head -1 | cut -d= -f2-) + if [ -n "$human_token" ]; then + export HUMAN_TOKEN="$human_token" + echo " Human token preserved (use --rotate-tokens to force)" + fi + fi + + if [ -z "$human_token" ]; then + # Delete existing human token if present (token sha1 is only returned at creation time) + local existing_human_token_id + existing_human_token_id=$(curl -sf \ + -u "${admin_user}:${admin_pass}" \ + "${forge_url}/api/v1/users/${human_user}/tokens" 2>/dev/null \ + | jq -r '.[] | select(.name == "disinto-human-token") | .id') || existing_human_token_id="" + if [ -n "$existing_human_token_id" ]; then + curl -sf -X DELETE \ + -u "${admin_user}:${admin_pass}" \ + "${forge_url}/api/v1/users/${human_user}/tokens/${existing_human_token_id}" >/dev/null 2>&1 || true + fi + + # Create human token (use admin_pass since human_user == admin_user) + human_token=$(curl -sf -X POST \ + -u "${admin_user}:${admin_pass}" \ + -H "Content-Type: application/json" \ + "${forge_url}/api/v1/users/${human_user}/tokens" \ + -d '{"name":"disinto-human-token","scopes":["all"]}' 2>/dev/null \ + | jq -r '.sha1 // empty') || human_token="" + + if [ -n "$human_token" ]; then + # Store human token in .env + if grep -q '^HUMAN_TOKEN=' "$env_file" 2>/dev/null; then + sed -i "s|^HUMAN_TOKEN=.*|HUMAN_TOKEN=${human_token}|" "$env_file" + else + printf 'HUMAN_TOKEN=%s\n' "$human_token" >> "$env_file" + fi + export HUMAN_TOKEN="$human_token" + echo " Human token generated and saved (HUMAN_TOKEN)" fi - export HUMAN_TOKEN="$human_token" - echo " Human token saved (HUMAN_TOKEN)" fi # Create bot users and tokens @@ -719,7 +747,7 @@ setup_forge() { fi # Add all bot users as collaborators with appropriate permissions - # dev-bot: write (PR creation via lib/vault.sh) + # dev-bot: write (PR creation via lib/action-vault.sh) # review-bot: read (PR review) # planner-bot: write (prerequisites.md, memory) # gardener-bot: write (backlog grooming) diff --git a/lib/generators.sh b/lib/generators.sh index 72f030e..02af667 100644 --- a/lib/generators.sh +++ b/lib/generators.sh @@ -100,9 +100,7 @@ _generate_local_model_services() { cat >> "$temp_file" <<EOF agents-${service_name}: - build: - context: . - dockerfile: docker/agents/Dockerfile + image: ghcr.io/disinto/agents:\${DISINTO_IMAGE_TAG:-latest} container_name: disinto-agents-${service_name} restart: unless-stopped security_opt: @@ -111,9 +109,9 @@ _generate_local_model_services() { - agents-${service_name}-data:/home/agent/data - project-repos:/home/agent/repos - \${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:\${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} - - \${HOME}/.claude.json:/home/agent/.claude.json:ro - - CLAUDE_BIN_PLACEHOLDER:/usr/local/bin/claude:ro - - \${HOME}/.ssh:/home/agent/.ssh:ro + - \${CLAUDE_CONFIG_FILE:-\${HOME}/.claude.json}:/home/agent/.claude.json:ro + - \${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - \${AGENT_SSH_DIR:-\${HOME}/.ssh}:/home/agent/.ssh:ro environment: FORGE_URL: http://forgejo:3000 FORGE_REPO: ${FORGE_REPO:-disinto-admin/disinto} @@ -142,6 +140,7 @@ _generate_local_model_services() { GARDENER_INTERVAL: "${GARDENER_INTERVAL:-21600}" ARCHITECT_INTERVAL: "${ARCHITECT_INTERVAL:-21600}" PLANNER_INTERVAL: "${PLANNER_INTERVAL:-43200}" + SUPERVISOR_INTERVAL: "${SUPERVISOR_INTERVAL:-1200}" depends_on: forgejo: condition: service_healthy @@ -233,6 +232,7 @@ for name, config in agents.items(): # to materialize a working stack on a fresh checkout. _generate_compose_impl() { local forge_port="${1:-3000}" + local use_build="${2:-false}" local compose_file="${FACTORY_ROOT}/docker-compose.yml" # Check if compose file already exists @@ -296,6 +296,7 @@ services: WOODPECKER_AGENT_SECRET: ${WOODPECKER_AGENT_SECRET:-} WOODPECKER_DATABASE_DRIVER: sqlite3 WOODPECKER_DATABASE_DATASOURCE: /var/lib/woodpecker/woodpecker.sqlite + WOODPECKER_PLUGINS_PRIVILEGED: ${WOODPECKER_PLUGINS_PRIVILEGED:-plugins/docker} WOODPECKER_ENVIRONMENT: "FORGE_TOKEN:${FORGE_TOKEN}" depends_on: forgejo: @@ -318,15 +319,19 @@ services: WOODPECKER_AGENT_SECRET: ${WOODPECKER_AGENT_SECRET:-} WOODPECKER_GRPC_SECURE: "false" WOODPECKER_HEALTHCHECK_ADDR: ":3333" - WOODPECKER_BACKEND_DOCKER_NETWORK: disinto_disinto-net + WOODPECKER_BACKEND_DOCKER_NETWORK: ${WOODPECKER_CI_NETWORK:-disinto_disinto-net} WOODPECKER_MAX_WORKFLOWS: 1 + healthcheck: + test: ["CMD", "wget", "-q", "--spider", "http://localhost:3333/healthz"] + interval: 30s + timeout: 5s + retries: 3 + start_period: 15s depends_on: - woodpecker agents: - build: - context: . - dockerfile: docker/agents/Dockerfile + image: ghcr.io/disinto/agents:${DISINTO_IMAGE_TAG:-latest} container_name: disinto-agents restart: unless-stopped security_opt: @@ -335,11 +340,14 @@ services: - agent-data:/home/agent/data - project-repos:/home/agent/repos - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} - - ${HOME}/.claude.json:/home/agent/.claude.json:ro - - CLAUDE_BIN_PLACEHOLDER:/usr/local/bin/claude:ro - - ${HOME}/.ssh:/home/agent/.ssh:ro - - ${HOME}/.config/sops/age:/home/agent/.config/sops/age:ro + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro + - ${SOPS_AGE_DIR:-${HOME}/.config/sops/age}:/home/agent/.config/sops/age:ro - woodpecker-data:/woodpecker-data:ro + - ./projects:/home/agent/disinto/projects:ro + - ./.env:/home/agent/disinto/.env:ro + - ./state:/home/agent/disinto/state environment: FORGE_URL: http://forgejo:3000 FORGE_REPO: ${FORGE_REPO:-disinto-admin/disinto} @@ -371,8 +379,14 @@ services: PLANNER_INTERVAL: ${PLANNER_INTERVAL:-43200} # IMPORTANT: agents get explicit environment variables (forge tokens, CI tokens, config). # Vault-only secrets (GITHUB_TOKEN, CLAWHUB_TOKEN, deploy keys) live in - # .env.vault.enc and are NEVER injected here — only the runner - # container receives them at fire time (AD-006, #745). + # secrets/*.enc and are NEVER injected here — only the runner + # container receives them at fire time (AD-006, #745, #777). + healthcheck: + test: ["CMD", "pgrep", "-f", "entrypoint.sh"] + interval: 60s + timeout: 5s + retries: 3 + start_period: 30s depends_on: forgejo: condition: service_healthy @@ -381,10 +395,137 @@ services: networks: - disinto-net - runner: +COMPOSEEOF + + # ── Conditional agents-llama block (ENABLE_LLAMA_AGENT=1) ────────────── + # Local-Qwen dev agent — gated on ENABLE_LLAMA_AGENT so factories without + # a local llama endpoint don't try to start it. See docs/agents-llama.md. + if [ "${ENABLE_LLAMA_AGENT:-0}" = "1" ]; then + cat >> "$compose_file" <<'LLAMAEOF' + + agents-llama: build: context: . dockerfile: docker/agents/Dockerfile + container_name: disinto-agents-llama + restart: unless-stopped + security_opt: + - apparmor=unconfined + volumes: + - agent-data:/home/agent/data + - project-repos:/home/agent/repos + - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro + - ${SOPS_AGE_DIR:-${HOME}/.config/sops/age}:/home/agent/.config/sops/age:ro + - woodpecker-data:/woodpecker-data:ro + environment: + FORGE_URL: http://forgejo:3000 + FORGE_REPO: ${FORGE_REPO:-disinto-admin/disinto} + FORGE_TOKEN: ${FORGE_TOKEN_LLAMA:-} + FORGE_PASS: ${FORGE_PASS_LLAMA:-} + FORGE_BOT_USERNAMES: ${FORGE_BOT_USERNAMES:-} + WOODPECKER_TOKEN: ${WOODPECKER_TOKEN:-} + CLAUDE_TIMEOUT: ${CLAUDE_TIMEOUT:-7200} + CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC: ${CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC:-1} + CLAUDE_AUTOCOMPACT_PCT_OVERRIDE: "60" + ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-} + ANTHROPIC_BASE_URL: ${ANTHROPIC_BASE_URL:-} + FORGE_ADMIN_PASS: ${FORGE_ADMIN_PASS:-} + DISINTO_CONTAINER: "1" + PROJECT_NAME: ${PROJECT_NAME:-project} + PROJECT_REPO_ROOT: /home/agent/repos/${PROJECT_NAME:-project} + WOODPECKER_DATA_DIR: /woodpecker-data + WOODPECKER_REPO_ID: "PLACEHOLDER_WP_REPO_ID" + CLAUDE_CONFIG_DIR: ${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config} + POLL_INTERVAL: ${POLL_INTERVAL:-300} + AGENT_ROLES: dev + healthcheck: + test: ["CMD", "pgrep", "-f", "entrypoint.sh"] + interval: 60s + timeout: 5s + retries: 3 + start_period: 30s + depends_on: + forgejo: + condition: service_healthy + networks: + - disinto-net + + agents-llama-all: + build: + context: . + dockerfile: docker/agents/Dockerfile + container_name: disinto-agents-llama-all + restart: unless-stopped + profiles: ["agents-llama-all"] + security_opt: + - apparmor=unconfined + volumes: + - agent-data:/home/agent/data + - project-repos:/home/agent/repos + - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro + - ${AGENT_SSH_DIR:-${HOME}/.ssh}:/home/agent/.ssh:ro + - ${SOPS_AGE_DIR:-${HOME}/.config/sops/age}:/home/agent/.config/sops/age:ro + - woodpecker-data:/woodpecker-data:ro + environment: + FORGE_URL: http://forgejo:3000 + FORGE_REPO: ${FORGE_REPO:-disinto-admin/disinto} + FORGE_TOKEN: ${FORGE_TOKEN_LLAMA:-} + FORGE_PASS: ${FORGE_PASS_LLAMA:-} + FORGE_REVIEW_TOKEN: ${FORGE_REVIEW_TOKEN:-} + FORGE_PLANNER_TOKEN: ${FORGE_PLANNER_TOKEN:-} + FORGE_GARDENER_TOKEN: ${FORGE_GARDENER_TOKEN:-} + FORGE_VAULT_TOKEN: ${FORGE_VAULT_TOKEN:-} + FORGE_SUPERVISOR_TOKEN: ${FORGE_SUPERVISOR_TOKEN:-} + FORGE_PREDICTOR_TOKEN: ${FORGE_PREDICTOR_TOKEN:-} + FORGE_ARCHITECT_TOKEN: ${FORGE_ARCHITECT_TOKEN:-} + FORGE_FILER_TOKEN: ${FORGE_FILER_TOKEN:-} + FORGE_BOT_USERNAMES: ${FORGE_BOT_USERNAMES:-} + WOODPECKER_TOKEN: ${WOODPECKER_TOKEN:-} + CLAUDE_TIMEOUT: ${CLAUDE_TIMEOUT:-7200} + CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC: ${CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC:-1} + CLAUDE_AUTOCOMPACT_PCT_OVERRIDE: "60" + CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS: "1" + ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-} + ANTHROPIC_BASE_URL: ${ANTHROPIC_BASE_URL:-} + FORGE_ADMIN_PASS: ${FORGE_ADMIN_PASS:-} + DISINTO_CONTAINER: "1" + PROJECT_NAME: ${PROJECT_NAME:-project} + PROJECT_REPO_ROOT: /home/agent/repos/${PROJECT_NAME:-project} + WOODPECKER_DATA_DIR: /woodpecker-data + WOODPECKER_REPO_ID: "PLACEHOLDER_WP_REPO_ID" + CLAUDE_CONFIG_DIR: ${CLAUDE_CONFIG_DIR:-/var/lib/disinto/claude-shared/config} + POLL_INTERVAL: ${POLL_INTERVAL:-300} + GARDENER_INTERVAL: ${GARDENER_INTERVAL:-21600} + ARCHITECT_INTERVAL: ${ARCHITECT_INTERVAL:-21600} + PLANNER_INTERVAL: ${PLANNER_INTERVAL:-43200} + SUPERVISOR_INTERVAL: ${SUPERVISOR_INTERVAL:-1200} + AGENT_ROLES: review,dev,gardener,architect,planner,predictor,supervisor + healthcheck: + test: ["CMD", "pgrep", "-f", "entrypoint.sh"] + interval: 60s + timeout: 5s + retries: 3 + start_period: 30s + depends_on: + forgejo: + condition: service_healthy + woodpecker: + condition: service_started + networks: + - disinto-net +LLAMAEOF + fi + + # Resume the rest of the compose file (runner onward) + cat >> "$compose_file" <<'COMPOSEEOF' + + runner: + image: ghcr.io/disinto/agents:${DISINTO_IMAGE_TAG:-latest} profiles: ["vault"] security_opt: - apparmor=unconfined @@ -405,8 +546,9 @@ services: # Edge proxy — reverse proxy to Forgejo, Woodpecker, and staging # Serves on ports 80/443, routes based on path edge: - build: ./docker/edge + image: ghcr.io/disinto/edge:${DISINTO_IMAGE_TAG:-latest} container_name: disinto-edge + restart: unless-stopped security_opt: - apparmor=unconfined ports: @@ -441,7 +583,13 @@ services: - /var/run/docker.sock:/var/run/docker.sock - ./secrets/tunnel_key:/run/secrets/tunnel_key:ro - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared} - - ${HOME}/.claude.json:/home/agent/.claude.json:ro + - ${CLAUDE_CONFIG_FILE:-${HOME}/.claude.json}:/home/agent/.claude.json:ro + healthcheck: + test: ["CMD", "curl", "-fsS", "http://localhost:2019/config/"] + interval: 30s + timeout: 5s + retries: 3 + start_period: 15s depends_on: forgejo: condition: service_healthy @@ -459,6 +607,12 @@ services: command: ["caddy", "file-server", "--root", "/srv/site"] security_opt: - apparmor=unconfined + healthcheck: + test: ["CMD", "wget", "-q", "--spider", "http://localhost:2019/config/"] + interval: 30s + timeout: 5s + retries: 3 + start_period: 10s volumes: - ./docker:/srv/site:ro networks: @@ -499,7 +653,7 @@ services: memswap_limit: 512m volumes: # Mount claude binary from host (same as agents) - - CLAUDE_BIN_PLACEHOLDER:/usr/local/bin/claude:ro + - ${CLAUDE_BIN_DIR}:/usr/local/bin/claude:ro # Throwaway named volume for chat config (isolated from host ~/.claude) - chat-config:/var/chat/config # Chat history persistence: per-user NDJSON files on bind-mounted host volume @@ -518,6 +672,12 @@ services: CHAT_MAX_REQUESTS_PER_HOUR: ${CHAT_MAX_REQUESTS_PER_HOUR:-60} CHAT_MAX_REQUESTS_PER_DAY: ${CHAT_MAX_REQUESTS_PER_DAY:-500} CHAT_MAX_TOKENS_PER_DAY: ${CHAT_MAX_TOKENS_PER_DAY:-1000000} + healthcheck: + test: ["CMD", "python3", "-c", "import urllib.request; urllib.request.urlopen('http://localhost:8080/health')"] + interval: 30s + timeout: 5s + retries: 3 + start_period: 10s networks: - disinto-net @@ -556,20 +716,35 @@ COMPOSEEOF fi # Append local-model agent services if any are configured - # (must run before CLAUDE_BIN_PLACEHOLDER substitution so the placeholder - # in local-model services is also resolved) _generate_local_model_services "$compose_file" - # Patch the Claude CLI binary path — resolve from host PATH at init time. + # Resolve the Claude CLI binary path and persist as CLAUDE_BIN_DIR in .env. + # docker-compose.yml references ${CLAUDE_BIN_DIR} so the value must be set. local claude_bin claude_bin="$(command -v claude 2>/dev/null || true)" if [ -n "$claude_bin" ]; then - # Resolve symlinks to get the real binary path claude_bin="$(readlink -f "$claude_bin")" - sed -i "s|CLAUDE_BIN_PLACEHOLDER|${claude_bin}|g" "$compose_file" else - echo "Warning: claude CLI not found in PATH — update docker-compose.yml volumes manually" >&2 - sed -i "s|CLAUDE_BIN_PLACEHOLDER|/usr/local/bin/claude|g" "$compose_file" + echo "Warning: claude CLI not found in PATH — set CLAUDE_BIN_DIR in .env manually" >&2 + claude_bin="/usr/local/bin/claude" + fi + # Persist CLAUDE_BIN_DIR into .env so docker-compose can resolve it. + local env_file="${FACTORY_ROOT}/.env" + if [ -f "$env_file" ]; then + if grep -q "^CLAUDE_BIN_DIR=" "$env_file" 2>/dev/null; then + sed -i "s|^CLAUDE_BIN_DIR=.*|CLAUDE_BIN_DIR=${claude_bin}|" "$env_file" + else + printf 'CLAUDE_BIN_DIR=%s\n' "$claude_bin" >> "$env_file" + fi + else + printf 'CLAUDE_BIN_DIR=%s\n' "$claude_bin" > "$env_file" + fi + + # In build mode, replace image: with build: for locally-built images + if [ "$use_build" = true ]; then + sed -i 's|^\( agents:\)|\1|' "$compose_file" + sed -i '/^ image: ghcr\.io\/disinto\/agents:/{s|image: ghcr\.io/disinto/agents:.*|build:\n context: .\n dockerfile: docker/agents/Dockerfile|}' "$compose_file" + sed -i '/^ image: ghcr\.io\/disinto\/edge:/{s|image: ghcr\.io/disinto/edge:.*|build: ./docker/edge|}' "$compose_file" fi echo "Created: ${compose_file}" @@ -588,7 +763,11 @@ _generate_agent_docker_impl() { fi } -# Generate docker/Caddyfile template for edge proxy. +# Generate docker/Caddyfile for the edge proxy. +# **CANONICAL SOURCE**: This generator is the single source of truth for the Caddyfile. +# Output path: ${FACTORY_ROOT}/docker/Caddyfile (gitignored — generated artifact). +# The edge compose service mounts this path as /etc/caddy/Caddyfile. +# On a fresh clone, `disinto init` calls generate_caddyfile before first `disinto up`. _generate_caddyfile_impl() { local docker_dir="${FACTORY_ROOT}/docker" local caddyfile="${docker_dir}/Caddyfile" diff --git a/lib/hvault.sh b/lib/hvault.sh new file mode 100644 index 0000000..b1e0d62 --- /dev/null +++ b/lib/hvault.sh @@ -0,0 +1,279 @@ +#!/usr/bin/env bash +# hvault.sh — HashiCorp Vault helper module +# +# Typed, audited helpers for Vault KV v2 access so no script re-implements +# `curl -H "X-Vault-Token: ..."` ad-hoc. +# +# Usage: source this file, then call any hvault_* function. +# +# Environment: +# VAULT_ADDR — Vault server address (required, no default) +# VAULT_TOKEN — auth token (precedence: env > /etc/vault.d/root.token) +# +# All functions emit structured JSON errors to stderr on failure. + +set -euo pipefail + +# ── Internal helpers ───────────────────────────────────────────────────────── + +# _hvault_err — emit structured JSON error to stderr +# Args: func_name, message, [detail] +_hvault_err() { + local func="$1" msg="$2" detail="${3:-}" + jq -n --arg func "$func" --arg msg "$msg" --arg detail "$detail" \ + '{error:true,function:$func,message:$msg,detail:$detail}' >&2 +} + +# _hvault_resolve_token — resolve VAULT_TOKEN from env or token file +_hvault_resolve_token() { + if [ -n "${VAULT_TOKEN:-}" ]; then + return 0 + fi + local token_file="/etc/vault.d/root.token" + if [ -f "$token_file" ]; then + VAULT_TOKEN="$(cat "$token_file")" + export VAULT_TOKEN + return 0 + fi + return 1 +} + +# _hvault_check_prereqs — validate VAULT_ADDR and VAULT_TOKEN are set +# Args: caller function name +_hvault_check_prereqs() { + local caller="$1" + if [ -z "${VAULT_ADDR:-}" ]; then + _hvault_err "$caller" "VAULT_ADDR is not set" "export VAULT_ADDR before calling $caller" + return 1 + fi + if ! _hvault_resolve_token; then + _hvault_err "$caller" "VAULT_TOKEN is not set and /etc/vault.d/root.token not found" \ + "export VAULT_TOKEN or write token to /etc/vault.d/root.token" + return 1 + fi +} + +# _hvault_request — execute a Vault API request +# Args: method, path, [data] +# Outputs: response body to stdout +# Returns: 0 on 2xx, 1 otherwise (error JSON to stderr) +_hvault_request() { + local method="$1" path="$2" data="${3:-}" + local url="${VAULT_ADDR}/v1/${path}" + local http_code body + local tmpfile + tmpfile="$(mktemp)" + + local curl_args=( + -s + -w '%{http_code}' + -H "X-Vault-Token: ${VAULT_TOKEN}" + -H "Content-Type: application/json" + -X "$method" + -o "$tmpfile" + ) + if [ -n "$data" ]; then + curl_args+=(-d "$data") + fi + + http_code="$(curl "${curl_args[@]}" "$url")" || { + _hvault_err "_hvault_request" "curl failed" "url=$url" + rm -f "$tmpfile" + return 1 + } + + body="$(cat "$tmpfile")" + rm -f "$tmpfile" + + # Check HTTP status — 2xx is success + case "$http_code" in + 2[0-9][0-9]) + printf '%s' "$body" + return 0 + ;; + *) + _hvault_err "_hvault_request" "HTTP $http_code" "$body" + return 1 + ;; + esac +} + +# ── Public API ─────────────────────────────────────────────────────────────── + +# hvault_kv_get PATH [KEY] +# Read a KV v2 secret at PATH, optionally extract a single KEY. +# Outputs: JSON value (full data object, or single key value) +hvault_kv_get() { + local path="${1:-}" + local key="${2:-}" + + if [ -z "$path" ]; then + _hvault_err "hvault_kv_get" "PATH is required" "usage: hvault_kv_get PATH [KEY]" + return 1 + fi + _hvault_check_prereqs "hvault_kv_get" || return 1 + + local response + response="$(_hvault_request GET "secret/data/${path}")" || return 1 + + if [ -n "$key" ]; then + printf '%s' "$response" | jq -e -r --arg key "$key" '.data.data[$key]' 2>/dev/null || { + _hvault_err "hvault_kv_get" "key not found" "key=$key path=$path" + return 1 + } + else + printf '%s' "$response" | jq -e '.data.data' 2>/dev/null || { + _hvault_err "hvault_kv_get" "failed to parse response" "path=$path" + return 1 + } + fi +} + +# hvault_kv_put PATH KEY=VAL [KEY=VAL ...] +# Write a KV v2 secret at PATH. Accepts one or more KEY=VAL pairs. +hvault_kv_put() { + local path="${1:-}" + shift || true + + if [ -z "$path" ] || [ $# -eq 0 ]; then + _hvault_err "hvault_kv_put" "PATH and at least one KEY=VAL required" \ + "usage: hvault_kv_put PATH KEY=VAL [KEY=VAL ...]" + return 1 + fi + _hvault_check_prereqs "hvault_kv_put" || return 1 + + # Build JSON payload from KEY=VAL pairs entirely via jq + local payload='{"data":{}}' + for kv in "$@"; do + local k="${kv%%=*}" + local v="${kv#*=}" + if [ "$k" = "$kv" ]; then + _hvault_err "hvault_kv_put" "invalid KEY=VAL pair" "got: $kv" + return 1 + fi + payload="$(printf '%s' "$payload" | jq --arg k "$k" --arg v "$v" '.data[$k] = $v')" + done + + _hvault_request POST "secret/data/${path}" "$payload" >/dev/null +} + +# hvault_kv_list PATH +# List keys at a KV v2 path. +# Outputs: JSON array of key names +hvault_kv_list() { + local path="${1:-}" + + if [ -z "$path" ]; then + _hvault_err "hvault_kv_list" "PATH is required" "usage: hvault_kv_list PATH" + return 1 + fi + _hvault_check_prereqs "hvault_kv_list" || return 1 + + local response + response="$(_hvault_request LIST "secret/metadata/${path}")" || return 1 + + printf '%s' "$response" | jq -e '.data.keys' 2>/dev/null || { + _hvault_err "hvault_kv_list" "failed to parse response" "path=$path" + return 1 + } +} + +# hvault_policy_apply NAME FILE +# Idempotent policy upsert — create or update a Vault policy. +hvault_policy_apply() { + local name="${1:-}" + local file="${2:-}" + + if [ -z "$name" ] || [ -z "$file" ]; then + _hvault_err "hvault_policy_apply" "NAME and FILE are required" \ + "usage: hvault_policy_apply NAME FILE" + return 1 + fi + if [ ! -f "$file" ]; then + _hvault_err "hvault_policy_apply" "policy file not found" "file=$file" + return 1 + fi + _hvault_check_prereqs "hvault_policy_apply" || return 1 + + local policy_content + policy_content="$(cat "$file")" + local payload + payload="$(jq -n --arg policy "$policy_content" '{"policy": $policy}')" + + _hvault_request PUT "sys/policies/acl/${name}" "$payload" >/dev/null +} + +# hvault_jwt_login ROLE JWT +# Exchange a JWT for a short-lived Vault token. +# Outputs: client token string +hvault_jwt_login() { + local role="${1:-}" + local jwt="${2:-}" + + if [ -z "$role" ] || [ -z "$jwt" ]; then + _hvault_err "hvault_jwt_login" "ROLE and JWT are required" \ + "usage: hvault_jwt_login ROLE JWT" + return 1 + fi + # Only need VAULT_ADDR, not VAULT_TOKEN (we're obtaining a token) + if [ -z "${VAULT_ADDR:-}" ]; then + _hvault_err "hvault_jwt_login" "VAULT_ADDR is not set" + return 1 + fi + + local payload + payload="$(jq -n --arg role "$role" --arg jwt "$jwt" \ + '{"role": $role, "jwt": $jwt}')" + + local response + # JWT login does not require an existing token — use curl directly + local tmpfile http_code + tmpfile="$(mktemp)" + http_code="$(curl -s -w '%{http_code}' \ + -H "Content-Type: application/json" \ + -X POST \ + -d "$payload" \ + -o "$tmpfile" \ + "${VAULT_ADDR}/v1/auth/jwt/login")" || { + _hvault_err "hvault_jwt_login" "curl failed" + rm -f "$tmpfile" + return 1 + } + + local body + body="$(cat "$tmpfile")" + rm -f "$tmpfile" + + case "$http_code" in + 2[0-9][0-9]) + printf '%s' "$body" | jq -e -r '.auth.client_token' 2>/dev/null || { + _hvault_err "hvault_jwt_login" "failed to extract client_token" "$body" + return 1 + } + ;; + *) + _hvault_err "hvault_jwt_login" "HTTP $http_code" "$body" + return 1 + ;; + esac +} + +# hvault_token_lookup +# Returns TTL, policies, and accessor for the current token. +# Outputs: JSON object with ttl, policies, accessor fields +hvault_token_lookup() { + _hvault_check_prereqs "hvault_token_lookup" || return 1 + + local response + response="$(_hvault_request GET "auth/token/lookup-self")" || return 1 + + printf '%s' "$response" | jq -e '{ + ttl: .data.ttl, + policies: .data.policies, + accessor: .data.accessor, + display_name: .data.display_name + }' 2>/dev/null || { + _hvault_err "hvault_token_lookup" "failed to parse token info" + return 1 + } +} diff --git a/lib/init/nomad/install.sh b/lib/init/nomad/install.sh new file mode 100755 index 0000000..43397fd --- /dev/null +++ b/lib/init/nomad/install.sh @@ -0,0 +1,118 @@ +#!/usr/bin/env bash +# ============================================================================= +# lib/init/nomad/install.sh — Idempotent apt install of HashiCorp Nomad +# +# Part of the Nomad+Vault migration (S0.2, issue #822). Installs the `nomad` +# binary from the HashiCorp apt repository. Does NOT install Vault — S0.3 +# owns that. Does NOT configure, start, or enable a systemd unit — +# lib/init/nomad/systemd-nomad.sh owns that. Does NOT wire this script into +# `disinto init` — S0.4 owns that. +# +# Idempotency contract: +# - Running twice back-to-back is a no-op once the target version is +# installed and the apt source is in place. +# - Adds the HashiCorp apt keyring only if it is absent. +# - Adds the HashiCorp apt sources list only if it is absent. +# - Skips `apt-get install` entirely when the installed version already +# matches ${NOMAD_VERSION}. +# +# Configuration: +# NOMAD_VERSION — pinned Nomad version (default: see below). The apt +# package name is versioned as "nomad=<version>-1". +# +# Usage: +# sudo NOMAD_VERSION=1.9.5 lib/init/nomad/install.sh +# +# Exit codes: +# 0 success (installed or already present) +# 1 precondition failure (not Debian/Ubuntu, missing tools, not root) +# ============================================================================= +set -euo pipefail + +# Pin to a specific Nomad 1.x release. Bump here, not at call sites. +NOMAD_VERSION="${NOMAD_VERSION:-1.9.5}" + +HASHICORP_KEYRING="/usr/share/keyrings/hashicorp-archive-keyring.gpg" +HASHICORP_SOURCES="/etc/apt/sources.list.d/hashicorp.list" +HASHICORP_GPG_URL="https://apt.releases.hashicorp.com/gpg" +HASHICORP_REPO_URL="https://apt.releases.hashicorp.com" + +log() { printf '[install-nomad] %s\n' "$*"; } +die() { printf '[install-nomad] ERROR: %s\n' "$*" >&2; exit 1; } + +# ── Preconditions ──────────────────────────────────────────────────────────── +if [ "$(id -u)" -ne 0 ]; then + die "must run as root (needs apt-get + /usr/share/keyrings write access)" +fi + +for bin in apt-get gpg curl lsb_release; do + command -v "$bin" >/dev/null 2>&1 \ + || die "required binary not found: ${bin}" +done + +CODENAME="$(lsb_release -cs)" +[ -n "$CODENAME" ] || die "lsb_release returned empty codename" + +# ── Fast-path: already at desired version? ─────────────────────────────────── +installed_version="" +if command -v nomad >/dev/null 2>&1; then + # `nomad version` prints e.g. "Nomad v1.9.5" on the first line. + installed_version="$(nomad version 2>/dev/null \ + | awk 'NR==1 {sub(/^v/, "", $2); print $2; exit}')" +fi + +if [ "$installed_version" = "$NOMAD_VERSION" ]; then + log "nomad ${NOMAD_VERSION} already installed — nothing to do" + exit 0 +fi + +# ── Ensure HashiCorp apt keyring ───────────────────────────────────────────── +if [ ! -f "$HASHICORP_KEYRING" ]; then + log "adding HashiCorp apt keyring → ${HASHICORP_KEYRING}" + tmpkey="$(mktemp)" + trap 'rm -f "$tmpkey"' EXIT + curl -fsSL "$HASHICORP_GPG_URL" -o "$tmpkey" \ + || die "failed to fetch HashiCorp GPG key from ${HASHICORP_GPG_URL}" + gpg --dearmor -o "$HASHICORP_KEYRING" < "$tmpkey" \ + || die "failed to dearmor HashiCorp GPG key" + chmod 0644 "$HASHICORP_KEYRING" + rm -f "$tmpkey" + trap - EXIT +else + log "HashiCorp apt keyring already present" +fi + +# ── Ensure HashiCorp apt sources list ──────────────────────────────────────── +desired_source="deb [signed-by=${HASHICORP_KEYRING}] ${HASHICORP_REPO_URL} ${CODENAME} main" +if [ ! -f "$HASHICORP_SOURCES" ] \ + || ! grep -qxF "$desired_source" "$HASHICORP_SOURCES"; then + log "writing HashiCorp apt sources list → ${HASHICORP_SOURCES}" + printf '%s\n' "$desired_source" > "$HASHICORP_SOURCES" + apt_update_needed=1 +else + log "HashiCorp apt sources list already present" + apt_update_needed=0 +fi + +# ── Install the pinned version ─────────────────────────────────────────────── +if [ "$apt_update_needed" -eq 1 ]; then + log "running apt-get update" + DEBIAN_FRONTEND=noninteractive apt-get update -qq \ + || die "apt-get update failed" +fi + +# HashiCorp apt packages use the "<version>-1" package-revision suffix. +pkg_spec="nomad=${NOMAD_VERSION}-1" +log "installing ${pkg_spec}" +DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \ + "$pkg_spec" \ + || die "apt-get install ${pkg_spec} failed" + +# ── Verify ─────────────────────────────────────────────────────────────────── +final_version="$(nomad version 2>/dev/null \ + | awk 'NR==1 {sub(/^v/, "", $2); print $2; exit}')" +if [ "$final_version" != "$NOMAD_VERSION" ]; then + die "post-install check: expected ${NOMAD_VERSION}, got '${final_version}'" +fi + +log "nomad ${NOMAD_VERSION} installed successfully" diff --git a/lib/init/nomad/systemd-nomad.sh b/lib/init/nomad/systemd-nomad.sh new file mode 100755 index 0000000..e9db191 --- /dev/null +++ b/lib/init/nomad/systemd-nomad.sh @@ -0,0 +1,130 @@ +#!/usr/bin/env bash +# ============================================================================= +# lib/init/nomad/systemd-nomad.sh — Idempotent systemd unit installer for Nomad +# +# Part of the Nomad+Vault migration (S0.2, issue #822). Writes +# /etc/systemd/system/nomad.service pointing at /etc/nomad.d/ and runs +# `systemctl enable nomad` WITHOUT starting the service — we don't launch +# the cluster until S0.4 wires everything together. +# +# Idempotency contract: +# - Existing unit file is NOT rewritten when on-disk content already +# matches the desired content (avoids spurious `daemon-reload`). +# - `systemctl enable` on an already-enabled unit is a no-op. +# - This script is safe to run unconditionally before every factory boot. +# +# Preconditions: +# - nomad binary installed (see lib/init/nomad/install.sh) +# - /etc/nomad.d/ will hold server.hcl / client.hcl (placed by S0.4) +# +# Usage: +# sudo lib/init/nomad/systemd-nomad.sh +# +# Exit codes: +# 0 success (unit installed + enabled, or already so) +# 1 precondition failure (not root, no systemctl, no nomad binary) +# ============================================================================= +set -euo pipefail + +UNIT_PATH="/etc/systemd/system/nomad.service" +NOMAD_CONFIG_DIR="/etc/nomad.d" +NOMAD_DATA_DIR="/var/lib/nomad" + +log() { printf '[systemd-nomad] %s\n' "$*"; } +die() { printf '[systemd-nomad] ERROR: %s\n' "$*" >&2; exit 1; } + +# ── Preconditions ──────────────────────────────────────────────────────────── +if [ "$(id -u)" -ne 0 ]; then + die "must run as root (needs write access to ${UNIT_PATH})" +fi + +command -v systemctl >/dev/null 2>&1 \ + || die "systemctl not found (systemd is required)" + +NOMAD_BIN="$(command -v nomad 2>/dev/null || true)" +[ -n "$NOMAD_BIN" ] \ + || die "nomad binary not found — run lib/init/nomad/install.sh first" + +# ── Desired unit content ───────────────────────────────────────────────────── +# Upstream-recommended baseline (https://developer.hashicorp.com/nomad/docs/install/production/deployment-guide) +# trimmed for a single-node combined server+client dev box. +# - Wants=/After= network-online: nomad must have networking up. +# - User/Group=root: the Docker driver needs root to talk to dockerd. +# - LimitNOFILE/LimitNPROC=infinity: avoid Nomad's startup warning. +# - KillSignal=SIGINT: triggers Nomad's graceful shutdown path. +# - Restart=on-failure with a bounded burst to avoid crash-loops eating the +# journal when /etc/nomad.d/ is mis-configured. +read -r -d '' DESIRED_UNIT <<EOF || true +[Unit] +Description=Nomad +Documentation=https://developer.hashicorp.com/nomad/docs +Wants=network-online.target +After=network-online.target + +# When Docker is present, ensure dockerd is up before nomad starts — the +# Docker task driver needs the daemon socket available at startup. +Wants=docker.service +After=docker.service + +[Service] +Type=notify +User=root +Group=root +ExecReload=/bin/kill -HUP \$MAINPID +ExecStart=${NOMAD_BIN} agent -config=${NOMAD_CONFIG_DIR} +KillMode=process +KillSignal=SIGINT +LimitNOFILE=infinity +LimitNPROC=infinity +Restart=on-failure +RestartSec=2 +StartLimitBurst=3 +StartLimitIntervalSec=10 +TasksMax=infinity +OOMScoreAdjust=-1000 + +[Install] +WantedBy=multi-user.target +EOF + +# ── Ensure config + data dirs exist ────────────────────────────────────────── +# We do not populate /etc/nomad.d/ here (that's S0.4). We do create the +# directory so `nomad agent -config=/etc/nomad.d` doesn't error if the unit +# is started before hcl files are dropped in. +for d in "$NOMAD_CONFIG_DIR" "$NOMAD_DATA_DIR"; do + if [ ! -d "$d" ]; then + log "creating ${d}" + install -d -m 0755 "$d" + fi +done + +# ── Install unit file only if content differs ──────────────────────────────── +needs_reload=0 +if [ ! -f "$UNIT_PATH" ] \ + || ! printf '%s\n' "$DESIRED_UNIT" | cmp -s - "$UNIT_PATH"; then + log "writing unit → ${UNIT_PATH}" + tmp="$(mktemp)" + trap 'rm -f "$tmp"' EXIT + printf '%s\n' "$DESIRED_UNIT" > "$tmp" + install -m 0644 -o root -g root "$tmp" "$UNIT_PATH" + rm -f "$tmp" + trap - EXIT + needs_reload=1 +else + log "unit file already up to date" +fi + +# ── Reload + enable ────────────────────────────────────────────────────────── +if [ "$needs_reload" -eq 1 ]; then + log "systemctl daemon-reload" + systemctl daemon-reload +fi + +if systemctl is-enabled --quiet nomad.service 2>/dev/null; then + log "nomad.service already enabled" +else + log "systemctl enable nomad" + systemctl enable nomad.service >/dev/null +fi + +log "done — unit installed and enabled (NOT started; S0.4 brings the cluster up)" diff --git a/lib/mirrors.sh b/lib/mirrors.sh index 3ba561d..9b135c4 100644 --- a/lib/mirrors.sh +++ b/lib/mirrors.sh @@ -1,8 +1,10 @@ #!/usr/bin/env bash -# mirrors.sh — Push primary branch + tags to configured mirror remotes. +# mirrors.sh — Mirror helpers: push to remotes + register pull mirrors via API. # # Usage: source lib/mirrors.sh; mirror_push +# source lib/mirrors.sh; mirror_pull_register <clone_url> <owner> <repo_name> [interval] # Requires: PROJECT_REPO_ROOT, PRIMARY_BRANCH, MIRROR_* vars from load-project.sh +# FORGE_API_BASE, FORGE_TOKEN for pull-mirror registration # shellcheck disable=SC2154 # globals set by load-project.sh / calling script @@ -37,3 +39,73 @@ mirror_push() { log "mirror: pushed to ${name} (pid $!)" done } + +# --------------------------------------------------------------------------- +# mirror_pull_register — register a Forgejo pull mirror via the /repos/migrate API. +# +# Creates a new repo as a pull mirror of an external source. Works against +# empty target repos (the repo is created by the API call itself). +# +# Usage: +# mirror_pull_register <clone_url> <owner> <repo_name> [interval] +# +# Args: +# clone_url — HTTPS URL of the source repo (e.g. https://codeberg.org/johba/disinto.git) +# owner — Forgejo org or user that will own the mirror repo +# repo_name — name of the new mirror repo on Forgejo +# interval — sync interval (default: "8h0m0s"; Forgejo duration format) +# +# Requires: +# FORGE_API_BASE, FORGE_TOKEN (from env.sh) +# +# Returns 0 on success, 1 on failure. Prints the new repo JSON to stdout. +# --------------------------------------------------------------------------- +mirror_pull_register() { + local clone_url="$1" + local owner="$2" + local repo_name="$3" + local interval="${4:-8h0m0s}" + + if [ -z "${FORGE_API_BASE:-}" ] || [ -z "${FORGE_TOKEN:-}" ]; then + echo "ERROR: FORGE_API_BASE and FORGE_TOKEN must be set" >&2 + return 1 + fi + + if [ -z "$clone_url" ] || [ -z "$owner" ] || [ -z "$repo_name" ]; then + echo "Usage: mirror_pull_register <clone_url> <owner> <repo_name> [interval]" >&2 + return 1 + fi + + local payload + payload=$(jq -n \ + --arg clone_addr "$clone_url" \ + --arg repo_name "$repo_name" \ + --arg repo_owner "$owner" \ + --arg interval "$interval" \ + '{ + clone_addr: $clone_addr, + repo_name: $repo_name, + repo_owner: $repo_owner, + mirror: true, + mirror_interval: $interval, + service: "git" + }') + + local http_code body + body=$(curl -s -w "\n%{http_code}" -X POST \ + -H "Authorization: token ${FORGE_TOKEN}" \ + -H "Content-Type: application/json" \ + "${FORGE_API_BASE}/repos/migrate" \ + -d "$payload") + + http_code=$(printf '%s' "$body" | tail -n1) + body=$(printf '%s' "$body" | sed '$d') + + if [ "$http_code" -ge 200 ] && [ "$http_code" -lt 300 ]; then + printf '%s\n' "$body" + return 0 + else + echo "ERROR: mirror_pull_register failed (HTTP ${http_code}): ${body}" >&2 + return 1 + fi +} diff --git a/lib/release.sh b/lib/release.sh index 9ddf2bd..b9a3978 100644 --- a/lib/release.sh +++ b/lib/release.sh @@ -18,8 +18,8 @@ # ============================================================================= set -euo pipefail -# Source vault.sh for _vault_log helper -source "${FACTORY_ROOT}/lib/vault.sh" +# Source action-vault.sh for _vault_log helper +source "${FACTORY_ROOT}/lib/action-vault.sh" # Assert required globals are set before using this module. _assert_release_globals() { diff --git a/nomad/client.hcl b/nomad/client.hcl new file mode 100644 index 0000000..b90d5c1 --- /dev/null +++ b/nomad/client.hcl @@ -0,0 +1,88 @@ +# ============================================================================= +# nomad/client.hcl — Docker driver + host_volume declarations +# +# Part of the Nomad+Vault migration (S0.2, issue #822). Deployed to +# /etc/nomad.d/client.hcl on the factory dev box alongside server.hcl. +# +# This file owns: Docker driver plugin config + host_volume pre-wiring. +# server.hcl owns: agent role, bind, ports, data_dir. +# +# NOTE: Nomad merges every *.hcl under -config=/etc/nomad.d, so declaring +# a second `client { ... }` block here augments (not replaces) the one in +# server.hcl. On a single-node setup this file could be inlined into +# server.hcl — the split is for readability, not semantics. +# +# host_volume declarations let Nomad jobspecs mount factory state by name +# (volume = "forgejo-data", etc.) without coupling host paths into jobspec +# HCL. Host paths under /srv/disinto/* are created out-of-band by the +# orchestrator (S0.4) before any job references them. +# ============================================================================= + +client { + # forgejo git server data (repos, avatars, attachments). + host_volume "forgejo-data" { + path = "/srv/disinto/forgejo-data" + read_only = false + } + + # woodpecker CI data (pipeline artifacts, sqlite db). + host_volume "woodpecker-data" { + path = "/srv/disinto/woodpecker-data" + read_only = false + } + + # agent runtime data (claude config, logs, phase files). + host_volume "agent-data" { + path = "/srv/disinto/agent-data" + read_only = false + } + + # per-project git clones and worktrees. + host_volume "project-repos" { + path = "/srv/disinto/project-repos" + read_only = false + } + + # caddy config + ACME state. + host_volume "caddy-data" { + path = "/srv/disinto/caddy-data" + read_only = false + } + + # disinto chat transcripts + attachments. + host_volume "chat-history" { + path = "/srv/disinto/chat-history" + read_only = false + } + + # ops repo clone (vault actions, sprint artifacts, knowledge). + host_volume "ops-repo" { + path = "/srv/disinto/ops-repo" + read_only = false + } +} + +# Docker task driver. `volumes.enabled = true` is required so jobspecs +# can mount host_volume declarations defined above. `allow_privileged` +# stays false — no factory workload needs privileged containers today, +# and flipping it is an audit-worthy change. +plugin "docker" { + config { + allow_privileged = false + + volumes { + enabled = true + } + + # Leave images behind when jobs stop, so short job churn doesn't thrash + # the image cache. Factory disk is not constrained; `docker system prune` + # is the escape hatch. + gc { + image = false + container = true + dangling_containers { + enabled = true + } + } + } +} diff --git a/nomad/server.hcl b/nomad/server.hcl new file mode 100644 index 0000000..27c8b9c --- /dev/null +++ b/nomad/server.hcl @@ -0,0 +1,53 @@ +# ============================================================================= +# nomad/server.hcl — Single-node combined server+client configuration +# +# Part of the Nomad+Vault migration (S0.2, issue #822). Deployed to +# /etc/nomad.d/server.hcl on the factory dev box alongside client.hcl. +# +# This file owns: agent role, ports, bind, data directory. +# client.hcl owns: Docker driver plugin config + host_volume declarations. +# +# NOTE: On single-node setups these two files could be merged into one +# (Nomad auto-merges every *.hcl under -config=/etc/nomad.d). The split is +# purely for readability — role/bind/port vs. plugin/volume wiring. +# +# This is a factory dev-box baseline — TLS, ACLs, gossip encryption, and +# consul/vault integration are deliberately absent and land in later steps. +# ============================================================================= + +data_dir = "/var/lib/nomad" +bind_addr = "127.0.0.1" +log_level = "INFO" + +# All Nomad agent traffic stays on localhost — the factory box does not +# federate with peers. Ports are the Nomad defaults, pinned here so that +# future changes to these numbers are a visible diff. +ports { + http = 4646 + rpc = 4647 + serf = 4648 +} + +# Single-node combined mode: this agent is both the only server and the +# only client. bootstrap_expect=1 makes the server quorum-of-one. +server { + enabled = true + bootstrap_expect = 1 +} + +client { + enabled = true +} + +# Advertise localhost to self to avoid surprises if the default IP +# autodetection picks a transient interface (e.g. docker0, wg0). +advertise { + http = "127.0.0.1" + rpc = "127.0.0.1" + serf = "127.0.0.1" +} + +# UI on by default — same bind as http, no TLS (localhost only). +ui { + enabled = true +} diff --git a/planner/AGENTS.md b/planner/AGENTS.md index 5168eb4..aa784f4 100644 --- a/planner/AGENTS.md +++ b/planner/AGENTS.md @@ -1,4 +1,4 @@ -<!-- last-reviewed: cdb0408466d875cdd893c712e3b3f0d0b909d4ef --> +<!-- last-reviewed: c363ee0aea2ae447daab28c2c850d6abefc8c6b5 --> # Planner Agent **Role**: Strategic planning using a Prerequisite Tree (Theory of Constraints), @@ -34,7 +34,9 @@ will then sections) and marks the prerequisite as blocked-on-vault in the tree. Deduplication: checks pending/ + approved/ + fired/ before creating. Phase 4 (journal-and-memory): write updated prerequisite tree + daily journal entry (committed to ops repo) and update `$OPS_REPO_ROOT/knowledge/planner-memory.md`. -Phase 5 (commit-ops): commit all ops repo changes, push directly. +Phase 5 (commit-ops): commit all ops repo changes to a `planner/run-YYYY-MM-DD` +branch, then create a PR and walk it to merge via review-bot (`pr_create` → +`pr_walk_to_merge`), mirroring the architect's ops flow. No direct push to main. AGENTS.md maintenance is handled by the Gardener. **Artifacts use `$OPS_REPO_ROOT`**: All planner artifacts (journal, @@ -55,7 +57,7 @@ nervous system component, not work. creates tmux session, injects formula prompt, monitors phase file, handles crash recovery, cleans up - `formulas/run-planner.toml` — Execution spec: six steps (preflight, prediction-triage, update-prerequisite-tree, file-at-constraints, - journal-and-memory, commit-and-pr) with `needs` dependencies. Claude + journal-and-memory, commit-ops-changes) with `needs` dependencies. Claude executes all steps in a single interactive session with tool access - `formulas/groom-backlog.toml` — Grooming formula for backlog triage and grooming. (Note: the planner no longer dispatches breakdown mode — complex diff --git a/planner/planner-run.sh b/planner/planner-run.sh index 6c5bcb2..c567427 100755 --- a/planner/planner-run.sh +++ b/planner/planner-run.sh @@ -10,7 +10,9 @@ # 2. Load formula (formulas/run-planner.toml) # 3. Context: VISION.md, AGENTS.md, ops:RESOURCES.md, structural graph, # planner memory, journal entries -# 4. agent_run(worktree, prompt) → Claude plans, may push knowledge updates +# 4. Create ops branch planner/run-YYYY-MM-DD for changes +# 5. agent_run(worktree, prompt) → Claude plans, commits to ops branch +# 6. If ops branch has commits: pr_create → pr_walk_to_merge (review-bot) # # Usage: # planner-run.sh [projects/disinto.toml] # project config (default: disinto) @@ -35,6 +37,10 @@ source "$FACTORY_ROOT/lib/worktree.sh" source "$FACTORY_ROOT/lib/guard.sh" # shellcheck source=../lib/agent-sdk.sh source "$FACTORY_ROOT/lib/agent-sdk.sh" +# shellcheck source=../lib/ci-helpers.sh +source "$FACTORY_ROOT/lib/ci-helpers.sh" +# shellcheck source=../lib/pr-lifecycle.sh +source "$FACTORY_ROOT/lib/pr-lifecycle.sh" LOG_FILE="${DISINTO_LOG_DIR}/planner/planner.log" # shellcheck disable=SC2034 # consumed by agent-sdk.sh @@ -146,12 +152,69 @@ ${PROMPT_FOOTER}" # ── Create worktree ────────────────────────────────────────────────────── formula_worktree_setup "$WORKTREE" +# ── Prepare ops branch for PR-based merge (#765) ──────────────────────── +PLANNER_OPS_BRANCH="planner/run-$(date -u +%Y-%m-%d)" +( + cd "$OPS_REPO_ROOT" + git fetch origin "${PRIMARY_BRANCH}" --quiet 2>/dev/null || true + git checkout "${PRIMARY_BRANCH}" --quiet 2>/dev/null || true + git pull --ff-only origin "${PRIMARY_BRANCH}" --quiet 2>/dev/null || true + # Create (or reset to) a fresh branch from PRIMARY_BRANCH + git checkout -B "$PLANNER_OPS_BRANCH" "origin/${PRIMARY_BRANCH}" --quiet 2>/dev/null || \ + git checkout -b "$PLANNER_OPS_BRANCH" --quiet 2>/dev/null || true +) +log "ops branch: ${PLANNER_OPS_BRANCH}" + # ── Run agent ───────────────────────────────────────────────────────────── export CLAUDE_MODEL="opus" agent_run --worktree "$WORKTREE" "$PROMPT" log "agent_run complete" +# ── PR lifecycle: create PR on ops repo and walk to merge (#765) ───────── +OPS_FORGE_API="${FORGE_API_BASE}/repos/${FORGE_OPS_REPO}" +ops_has_commits=false +if ! git -C "$OPS_REPO_ROOT" diff --quiet "origin/${PRIMARY_BRANCH}..${PLANNER_OPS_BRANCH}" 2>/dev/null; then + ops_has_commits=true +fi + +if [ "$ops_has_commits" = "true" ]; then + log "ops branch has commits — creating PR" + # Push the branch to the ops remote + git -C "$OPS_REPO_ROOT" push origin "$PLANNER_OPS_BRANCH" --quiet 2>/dev/null || \ + git -C "$OPS_REPO_ROOT" push --force-with-lease origin "$PLANNER_OPS_BRANCH" 2>/dev/null + + # Temporarily point FORGE_API at the ops repo for pr-lifecycle functions + ORIG_FORGE_API="$FORGE_API" + export FORGE_API="$OPS_FORGE_API" + # Ops repo typically has no Woodpecker CI — skip CI polling + ORIG_WOODPECKER_REPO_ID="${WOODPECKER_REPO_ID:-2}" + export WOODPECKER_REPO_ID="0" + + PR_NUM=$(pr_create "$PLANNER_OPS_BRANCH" \ + "chore: planner run $(date -u +%Y-%m-%d)" \ + "Automated planner run — updates prerequisite tree, memory, and vault items." \ + "${PRIMARY_BRANCH}" \ + "$OPS_FORGE_API") || true + + if [ -n "$PR_NUM" ]; then + log "ops PR #${PR_NUM} created — walking to merge" + SESSION_ID=$(cat "$SID_FILE" 2>/dev/null || echo "planner-$$") + pr_walk_to_merge "$PR_NUM" "$SESSION_ID" "$OPS_REPO_ROOT" 1 2 || { + log "ops PR #${PR_NUM} walk finished: ${_PR_WALK_EXIT_REASON:-unknown}" + } + log "ops PR #${PR_NUM} result: ${_PR_WALK_EXIT_REASON:-unknown}" + else + log "WARNING: failed to create ops PR for branch ${PLANNER_OPS_BRANCH}" + fi + + # Restore original FORGE_API + export FORGE_API="$ORIG_FORGE_API" + export WOODPECKER_REPO_ID="$ORIG_WOODPECKER_REPO_ID" +else + log "no ops changes — skipping PR creation" +fi + # Persist watermarks so next run can skip if nothing changed mkdir -p "$FACTORY_ROOT/state" echo "$CURRENT_SHA" > "$LAST_SHA_FILE" diff --git a/predictor/AGENTS.md b/predictor/AGENTS.md index b5391fe..c10e1f8 100644 --- a/predictor/AGENTS.md +++ b/predictor/AGENTS.md @@ -1,4 +1,4 @@ -<!-- last-reviewed: cdb0408466d875cdd893c712e3b3f0d0b909d4ef --> +<!-- last-reviewed: c363ee0aea2ae447daab28c2c850d6abefc8c6b5 --> # Predictor Agent **Role**: Abstract adversary (the "goblin"). Runs a 2-step formula diff --git a/review/AGENTS.md b/review/AGENTS.md index a3eb4c5..5137302 100644 --- a/review/AGENTS.md +++ b/review/AGENTS.md @@ -1,4 +1,4 @@ -<!-- last-reviewed: cdb0408466d875cdd893c712e3b3f0d0b909d4ef --> +<!-- last-reviewed: c363ee0aea2ae447daab28c2c850d6abefc8c6b5 --> # Review Agent **Role**: AI-powered PR review — post structured findings and formal diff --git a/supervisor/AGENTS.md b/supervisor/AGENTS.md index 46d7335..ef36ccb 100644 --- a/supervisor/AGENTS.md +++ b/supervisor/AGENTS.md @@ -1,4 +1,4 @@ -<!-- last-reviewed: cdb0408466d875cdd893c712e3b3f0d0b909d4ef --> +<!-- last-reviewed: c363ee0aea2ae447daab28c2c850d6abefc8c6b5 --> # Supervisor Agent **Role**: Health monitoring and auto-remediation, executed as a formula-driven @@ -7,13 +7,11 @@ then runs an interactive Claude session (sonnet) that assesses health, auto-fixe issues, and writes a daily journal. When blocked on external resources or human decisions, files vault items instead of escalating directly. -**Trigger**: `supervisor-run.sh` is invoked by the polling loop in `docker/edge/entrypoint-edge.sh` -every 20 minutes (line 50-53). Sources `lib/guard.sh` and calls `check_active supervisor` first -— skips if `$FACTORY_ROOT/state/.supervisor-active` is absent. Then runs `claude -p` via -`agent-sdk.sh`, injects `formulas/run-supervisor.toml` with pre-collected metrics as context, -and cleans up on completion or timeout (20 min max session). Note: the supervisor runs in the -**edge container** (`entrypoint-edge.sh`), not the agent container — this distinction matters -for operators debugging the factory. +**Trigger**: `supervisor-run.sh` is invoked by two polling loops: +- **Agents container** (`docker/agents/entrypoint.sh`): every `SUPERVISOR_INTERVAL` seconds (default 1200 = 20 min). Controlled by the `supervisor` role in `AGENT_ROLES` (included in the default seven-role set since P1/#801). Logs to `supervisor.log` in the agents container. +- **Edge container** (`docker/edge/entrypoint-edge.sh`): separate loop in the edge container (line 169-172). Runs independently of the agents container's polling schedule. + +Both invoke the same `supervisor-run.sh`. Sources `lib/guard.sh` and calls `check_active supervisor` first — skips if `$FACTORY_ROOT/state/.supervisor-active` is absent. Then runs `claude -p` via `agent-sdk.sh`, injects `formulas/run-supervisor.toml` with pre-collected metrics as context, and cleans up on completion or timeout. **Key files**: - `supervisor/supervisor-run.sh` — Polling loop participant + orchestrator: lock, memory guard, @@ -39,6 +37,7 @@ P3 (degraded PRs, circular deps, stale deps), P4 (housekeeping). **Environment variables consumed**: - `FORGE_TOKEN`, `FORGE_SUPERVISOR_TOKEN` (falls back to FORGE_TOKEN), `FORGE_REPO`, `FORGE_API`, `PROJECT_NAME`, `PROJECT_REPO_ROOT`, `OPS_REPO_ROOT` - `PRIMARY_BRANCH`, `CLAUDE_MODEL` (set to sonnet by supervisor-run.sh) +- `SUPERVISOR_INTERVAL` — polling interval in seconds for agents container (default 1200 = 20 min) - `WOODPECKER_TOKEN`, `WOODPECKER_SERVER`, `WOODPECKER_DB_PASSWORD`, `WOODPECKER_DB_USER`, `WOODPECKER_DB_HOST`, `WOODPECKER_DB_NAME` — CI database queries **Degraded mode (Issue #544)**: When `OPS_REPO_ROOT` is not set or the directory doesn't exist, the supervisor runs in degraded mode: diff --git a/tests/lib-hvault.bats b/tests/lib-hvault.bats new file mode 100644 index 0000000..628bc99 --- /dev/null +++ b/tests/lib-hvault.bats @@ -0,0 +1,215 @@ +#!/usr/bin/env bats +# tests/lib-hvault.bats — Unit tests for lib/hvault.sh +# +# Runs against a dev-mode Vault server (single binary, no LXC needed). +# CI launches vault server -dev inline before running these tests. + +VAULT_BIN="${VAULT_BIN:-vault}" + +setup_file() { + export TEST_DIR + TEST_DIR="$(cd "$(dirname "$BATS_TEST_FILENAME")/.." && pwd)" + + # Start dev-mode vault on a random port + export VAULT_DEV_PORT + VAULT_DEV_PORT="$(shuf -i 18200-18299 -n 1)" + export VAULT_ADDR="http://127.0.0.1:${VAULT_DEV_PORT}" + + "$VAULT_BIN" server -dev \ + -dev-listen-address="127.0.0.1:${VAULT_DEV_PORT}" \ + -dev-root-token-id="test-root-token" \ + -dev-no-store-token \ + &>"${BATS_FILE_TMPDIR}/vault.log" & + export VAULT_PID=$! + + export VAULT_TOKEN="test-root-token" + + # Wait for vault to be ready (up to 10s) + local i=0 + while ! curl -sf "${VAULT_ADDR}/v1/sys/health" >/dev/null 2>&1; do + sleep 0.5 + i=$((i + 1)) + if [ "$i" -ge 20 ]; then + echo "Vault failed to start. Log:" >&2 + cat "${BATS_FILE_TMPDIR}/vault.log" >&2 + return 1 + fi + done +} + +teardown_file() { + if [ -n "${VAULT_PID:-}" ]; then + kill "$VAULT_PID" 2>/dev/null || true + wait "$VAULT_PID" 2>/dev/null || true + fi +} + +setup() { + # Source the module under test + source "${TEST_DIR}/lib/hvault.sh" + export VAULT_ADDR VAULT_TOKEN +} + +# ── hvault_kv_put + hvault_kv_get ──────────────────────────────────────────── + +@test "hvault_kv_put writes and hvault_kv_get reads a secret" { + run hvault_kv_put "test/myapp" "username=admin" "password=s3cret" + [ "$status" -eq 0 ] + + run hvault_kv_get "test/myapp" + [ "$status" -eq 0 ] + echo "$output" | jq -e '.username == "admin"' + echo "$output" | jq -e '.password == "s3cret"' +} + +@test "hvault_kv_get extracts a single key" { + hvault_kv_put "test/single" "foo=bar" "baz=qux" + + run hvault_kv_get "test/single" "foo" + [ "$status" -eq 0 ] + [ "$output" = "bar" ] +} + +@test "hvault_kv_get fails for missing key" { + hvault_kv_put "test/keymiss" "exists=yes" + + run hvault_kv_get "test/keymiss" "nope" + [ "$status" -ne 0 ] +} + +@test "hvault_kv_get fails for missing path" { + run hvault_kv_get "test/does-not-exist-$(date +%s)" + [ "$status" -ne 0 ] +} + +@test "hvault_kv_put fails without KEY=VAL" { + run hvault_kv_put "test/bad" + [ "$status" -ne 0 ] + echo "$output" | grep -q '"error":true' || echo "$stderr" | grep -q '"error":true' +} + +@test "hvault_kv_put rejects malformed pair (no =)" { + run hvault_kv_put "test/bad2" "noequals" + [ "$status" -ne 0 ] +} + +@test "hvault_kv_get fails without PATH" { + run hvault_kv_get + [ "$status" -ne 0 ] +} + +# ── hvault_kv_list ─────────────────────────────────────────────────────────── + +@test "hvault_kv_list lists keys at a path" { + hvault_kv_put "test/listdir/a" "k=1" + hvault_kv_put "test/listdir/b" "k=2" + + run hvault_kv_list "test/listdir" + [ "$status" -eq 0 ] + echo "$output" | jq -e '. | length >= 2' + echo "$output" | jq -e 'index("a")' + echo "$output" | jq -e 'index("b")' +} + +@test "hvault_kv_list fails on nonexistent path" { + run hvault_kv_list "test/no-such-path-$(date +%s)" + [ "$status" -ne 0 ] +} + +@test "hvault_kv_list fails without PATH" { + run hvault_kv_list + [ "$status" -ne 0 ] +} + +# ── hvault_policy_apply ────────────────────────────────────────────────────── + +@test "hvault_policy_apply creates a policy" { + local pfile="${BATS_TEST_TMPDIR}/test-policy.hcl" + cat > "$pfile" <<'HCL' +path "secret/data/test/*" { + capabilities = ["read"] +} +HCL + + run hvault_policy_apply "test-reader" "$pfile" + [ "$status" -eq 0 ] + + # Verify the policy exists via Vault API + run curl -sf -H "X-Vault-Token: ${VAULT_TOKEN}" \ + "${VAULT_ADDR}/v1/sys/policies/acl/test-reader" + [ "$status" -eq 0 ] + echo "$output" | jq -e '.data.policy' | grep -q "secret/data/test" +} + +@test "hvault_policy_apply is idempotent" { + local pfile="${BATS_TEST_TMPDIR}/idem-policy.hcl" + printf 'path "secret/*" { capabilities = ["list"] }\n' > "$pfile" + + run hvault_policy_apply "idem-policy" "$pfile" + [ "$status" -eq 0 ] + + # Apply again — should succeed + run hvault_policy_apply "idem-policy" "$pfile" + [ "$status" -eq 0 ] +} + +@test "hvault_policy_apply fails with missing file" { + run hvault_policy_apply "bad-policy" "/nonexistent/policy.hcl" + [ "$status" -ne 0 ] +} + +@test "hvault_policy_apply fails without args" { + run hvault_policy_apply + [ "$status" -ne 0 ] +} + +# ── hvault_token_lookup ────────────────────────────────────────────────────── + +@test "hvault_token_lookup returns token info" { + run hvault_token_lookup + [ "$status" -eq 0 ] + echo "$output" | jq -e '.policies' + echo "$output" | jq -e '.accessor' + echo "$output" | jq -e 'has("ttl")' +} + +@test "hvault_token_lookup fails without VAULT_TOKEN" { + unset VAULT_TOKEN + run hvault_token_lookup + [ "$status" -ne 0 ] +} + +@test "hvault_token_lookup fails without VAULT_ADDR" { + unset VAULT_ADDR + run hvault_token_lookup + [ "$status" -ne 0 ] +} + +# ── hvault_jwt_login ───────────────────────────────────────────────────────── + +@test "hvault_jwt_login fails without VAULT_ADDR" { + unset VAULT_ADDR + run hvault_jwt_login "myrole" "fakejwt" + [ "$status" -ne 0 ] +} + +@test "hvault_jwt_login fails without args" { + run hvault_jwt_login + [ "$status" -ne 0 ] +} + +@test "hvault_jwt_login returns error for unconfigured jwt auth" { + # JWT auth backend is not enabled in dev mode by default — expect failure + run hvault_jwt_login "myrole" "eyJhbGciOiJSUzI1NiJ9.fake.sig" + [ "$status" -ne 0 ] +} + +# ── Env / prereq errors ───────────────────────────────────────────────────── + +@test "all functions fail with structured JSON error when VAULT_ADDR unset" { + unset VAULT_ADDR + for fn in hvault_kv_get hvault_kv_put hvault_kv_list hvault_policy_apply hvault_token_lookup; do + run $fn "dummy" "dummy" + [ "$status" -ne 0 ] + done +} diff --git a/tests/smoke-init.sh b/tests/smoke-init.sh index e8cd245..306f7ee 100644 --- a/tests/smoke-init.sh +++ b/tests/smoke-init.sh @@ -29,7 +29,8 @@ cleanup() { pkill -f "mock-forgejo.py" 2>/dev/null || true rm -rf "$MOCK_BIN" /tmp/smoke-test-repo \ "${FACTORY_ROOT}/projects/smoke-repo.toml" \ - /tmp/smoke-claude-shared /tmp/smoke-home-claude + /tmp/smoke-claude-shared /tmp/smoke-home-claude \ + /tmp/smoke-env-before-rerun /tmp/smoke-env-before-dryrun # Restore .env only if we created the backup if [ -f "${FACTORY_ROOT}/.env.smoke-backup" ]; then mv "${FACTORY_ROOT}/.env.smoke-backup" "${FACTORY_ROOT}/.env" @@ -178,8 +179,30 @@ else fail "disinto init exited non-zero" fi -# ── Idempotency test: run init again ─────────────────────────────────────── +# ── Dry-run test: must not modify state ──────────────────────────────────── +echo "=== Dry-run test ===" +cp "${FACTORY_ROOT}/.env" /tmp/smoke-env-before-dryrun +if bash "${FACTORY_ROOT}/bin/disinto" init \ + "${TEST_SLUG}" \ + --bare --yes --dry-run \ + --forge-url "$FORGE_URL" \ + --repo-root "/tmp/smoke-test-repo" 2>&1 | grep -q "Dry run complete"; then + pass "disinto init --dry-run exited successfully" +else + fail "disinto init --dry-run did not complete" +fi + +# Verify --dry-run did not modify .env +if diff -q /tmp/smoke-env-before-dryrun "${FACTORY_ROOT}/.env" >/dev/null 2>&1; then + pass "dry-run: .env unchanged" +else + fail "dry-run: .env was modified (should be read-only)" +fi +rm -f /tmp/smoke-env-before-dryrun + +# ── Idempotency test: run init again, verify .env is stable ──────────────── echo "=== Idempotency test: running disinto init again ===" +cp "${FACTORY_ROOT}/.env" /tmp/smoke-env-before-rerun if bash "${FACTORY_ROOT}/bin/disinto" init \ "${TEST_SLUG}" \ --bare --yes \ @@ -190,6 +213,29 @@ else fail "disinto init (re-run) exited non-zero" fi +# Verify .env is stable across re-runs (no token churn) +if diff -q /tmp/smoke-env-before-rerun "${FACTORY_ROOT}/.env" >/dev/null 2>&1; then + pass "idempotency: .env unchanged on re-run" +else + fail "idempotency: .env changed on re-run (token churn detected)" + diff /tmp/smoke-env-before-rerun "${FACTORY_ROOT}/.env" >&2 || true +fi +rm -f /tmp/smoke-env-before-rerun + +# Verify FORGE_ADMIN_TOKEN is stored in .env +if grep -q '^FORGE_ADMIN_TOKEN=' "${FACTORY_ROOT}/.env"; then + pass ".env contains FORGE_ADMIN_TOKEN" +else + fail ".env missing FORGE_ADMIN_TOKEN" +fi + +# Verify HUMAN_TOKEN is stored in .env +if grep -q '^HUMAN_TOKEN=' "${FACTORY_ROOT}/.env"; then + pass ".env contains HUMAN_TOKEN" +else + fail ".env missing HUMAN_TOKEN" +fi + # ── 4. Verify Forgejo state ───────────────────────────────────────────────── echo "=== 4/6 Verifying Forgejo state ===" diff --git a/tests/smoke-load-secret.sh b/tests/smoke-load-secret.sh new file mode 100644 index 0000000..2c409fa --- /dev/null +++ b/tests/smoke-load-secret.sh @@ -0,0 +1,162 @@ +#!/usr/bin/env bash +# tests/smoke-load-secret.sh — Unit tests for load_secret() precedence chain +# +# Covers the 4 precedence cases: +# 1. /secrets/<NAME>.env (Nomad template) +# 2. Current environment +# 3. secrets/<NAME>.enc (age-encrypted per-key file) +# 4. Default / empty fallback +# +# Required tools: bash, age (for case 3) + +set -euo pipefail + +FACTORY_ROOT="$(cd "$(dirname "$0")/.." && pwd)" + +fail() { printf 'FAIL: %s\n' "$*" >&2; FAILED=1; } +pass() { printf 'PASS: %s\n' "$*"; } +FAILED=0 + +# Set up a temp workspace and fake HOME so age key paths work +test_dir=$(mktemp -d) +fake_home=$(mktemp -d) +trap 'rm -rf "$test_dir" "$fake_home"' EXIT + +# Minimal env for sourcing env.sh's load_secret function without the full boot +# We source the function definition directly to isolate the unit under test. +# shellcheck disable=SC2034 +export USER="${USER:-test}" +export HOME="$fake_home" + +# Source env.sh to get load_secret (and FACTORY_ROOT) +source "${FACTORY_ROOT}/lib/env.sh" + +# ── Case 4: Default / empty fallback ──────────────────────────────────────── +echo "=== 1/5 Case 4: default fallback ===" + +unset TEST_SECRET_FALLBACK 2>/dev/null || true +val=$(load_secret TEST_SECRET_FALLBACK "my-default") +if [ "$val" = "my-default" ]; then + pass "load_secret returns default when nothing is set" +else + fail "Expected 'my-default', got '${val}'" +fi + +val=$(load_secret TEST_SECRET_FALLBACK) +if [ -z "$val" ]; then + pass "load_secret returns empty when no default and nothing set" +else + fail "Expected empty, got '${val}'" +fi + +# ── Case 2: Environment variable already set ──────────────────────────────── +echo "=== 2/5 Case 2: environment variable ===" + +export TEST_SECRET_ENV="from-environment" +val=$(load_secret TEST_SECRET_ENV "ignored-default") +if [ "$val" = "from-environment" ]; then + pass "load_secret returns env value over default" +else + fail "Expected 'from-environment', got '${val}'" +fi +unset TEST_SECRET_ENV + +# ── Case 3: Age-encrypted per-key file ────────────────────────────────────── +echo "=== 3/5 Case 3: age-encrypted secret ===" + +if command -v age &>/dev/null && command -v age-keygen &>/dev/null; then + # Generate a test age key + age_key_dir="${fake_home}/.config/sops/age" + mkdir -p "$age_key_dir" + age-keygen -o "${age_key_dir}/keys.txt" 2>/dev/null + pub_key=$(age-keygen -y "${age_key_dir}/keys.txt") + + # Create encrypted secret + secrets_dir="${FACTORY_ROOT}/secrets" + mkdir -p "$secrets_dir" + printf 'age-test-value' | age -r "$pub_key" -o "${secrets_dir}/TEST_SECRET_AGE.enc" + + unset TEST_SECRET_AGE 2>/dev/null || true + val=$(load_secret TEST_SECRET_AGE "fallback") + if [ "$val" = "age-test-value" ]; then + pass "load_secret decrypts age-encrypted secret" + else + fail "Expected 'age-test-value', got '${val}'" + fi + + # Verify caching: call load_secret directly (not in subshell) so export propagates + unset TEST_SECRET_AGE 2>/dev/null || true + load_secret TEST_SECRET_AGE >/dev/null + if [ "${TEST_SECRET_AGE:-}" = "age-test-value" ]; then + pass "load_secret caches decrypted value in environment (direct call)" + else + fail "Decrypted value not cached in environment" + fi + + # Clean up test secret + rm -f "${secrets_dir}/TEST_SECRET_AGE.enc" + rmdir "$secrets_dir" 2>/dev/null || true + unset TEST_SECRET_AGE +else + echo "SKIP: age/age-keygen not found — skipping age decryption test" +fi + +# ── Case 1: Nomad template path ──────────────────────────────────────────── +echo "=== 4/5 Case 1: Nomad template (/secrets/<NAME>.env) ===" + +nomad_dir="/secrets" +if [ -w "$(dirname "$nomad_dir")" ] 2>/dev/null || [ -w "$nomad_dir" ] 2>/dev/null; then + mkdir -p "$nomad_dir" + printf 'TEST_SECRET_NOMAD=from-nomad-template\n' > "${nomad_dir}/TEST_SECRET_NOMAD.env" + + # Even with env set, Nomad path takes precedence + export TEST_SECRET_NOMAD="from-env-should-lose" + val=$(load_secret TEST_SECRET_NOMAD "default") + if [ "$val" = "from-nomad-template" ]; then + pass "load_secret prefers Nomad template over env" + else + fail "Expected 'from-nomad-template', got '${val}'" + fi + + rm -f "${nomad_dir}/TEST_SECRET_NOMAD.env" + rmdir "$nomad_dir" 2>/dev/null || true + unset TEST_SECRET_NOMAD +else + echo "SKIP: /secrets not writable — skipping Nomad template test (needs root or container)" +fi + +# ── Precedence: env beats age ──────────────────────────────────────────── +echo "=== 5/5 Precedence: env beats age-encrypted ===" + +if command -v age &>/dev/null && command -v age-keygen &>/dev/null; then + age_key_dir="${fake_home}/.config/sops/age" + mkdir -p "$age_key_dir" + [ -f "${age_key_dir}/keys.txt" ] || age-keygen -o "${age_key_dir}/keys.txt" 2>/dev/null + pub_key=$(age-keygen -y "${age_key_dir}/keys.txt") + + secrets_dir="${FACTORY_ROOT}/secrets" + mkdir -p "$secrets_dir" + printf 'age-value-should-lose' | age -r "$pub_key" -o "${secrets_dir}/TEST_SECRET_PREC.enc" + + export TEST_SECRET_PREC="env-value-wins" + val=$(load_secret TEST_SECRET_PREC "default") + if [ "$val" = "env-value-wins" ]; then + pass "load_secret prefers env over age-encrypted file" + else + fail "Expected 'env-value-wins', got '${val}'" + fi + + rm -f "${secrets_dir}/TEST_SECRET_PREC.enc" + rmdir "$secrets_dir" 2>/dev/null || true + unset TEST_SECRET_PREC +else + echo "SKIP: age not found — skipping precedence test" +fi + +# ── Summary ─────────────────────────────────────────────────────────────── +echo "" +if [ "$FAILED" -ne 0 ]; then + echo "=== SMOKE-LOAD-SECRET TEST FAILED ===" + exit 1 +fi +echo "=== SMOKE-LOAD-SECRET TEST PASSED ===" diff --git a/tools/edge-control/README.md b/tools/edge-control/README.md index c49e78a..019b385 100644 --- a/tools/edge-control/README.md +++ b/tools/edge-control/README.md @@ -83,9 +83,12 @@ curl -sL https://raw.githubusercontent.com/disinto-admin/disinto/fix/issue-621/t - Permissions: `root:disinto-register 0750` 3. **Installs Caddy**: + - Backs up any pre-existing `/etc/caddy/Caddyfile` to `/etc/caddy/Caddyfile.pre-disinto` - Download Caddy with Gandi DNS plugin - Enable admin API on `127.0.0.1:2019` - Configure wildcard cert for `*.disinto.ai` via DNS-01 + - Creates `/etc/caddy/extra.d/` for operator-owned site blocks + - Emitted Caddyfile ends with `import /etc/caddy/extra.d/*.caddy` 4. **Sets up SSH**: - Creates `disinto-register` authorized_keys with forced command @@ -95,6 +98,27 @@ curl -sL https://raw.githubusercontent.com/disinto-admin/disinto/fix/issue-621/t - `/opt/disinto-edge/register.sh` — forced command handler - `/opt/disinto-edge/lib/*.sh` — helper libraries +## Operator-Owned Site Blocks + +Edge-control owns the top-level `/etc/caddy/Caddyfile` and dynamic `<project>.<DOMAIN_SUFFIX>` routes injected via the Caddy admin API. Operators own everything under `/etc/caddy/extra.d/`. + +To serve non-tunnel content (apex domain, www redirect, static sites), drop `.caddy` files into `/etc/caddy/extra.d/`: + +```bash +# Example: /etc/caddy/extra.d/landing.caddy +disinto.ai { + root * /home/debian/disinto-site + file_server +} + +# Example: /etc/caddy/extra.d/www-redirect.caddy +www.disinto.ai { + redir https://disinto.ai{uri} permanent +} +``` + +These files survive across `install.sh` re-runs. The `--extra-caddyfile <path>` flag overrides the default import glob (`/etc/caddy/extra.d/*.caddy`) if needed. + ## Usage ### Register a Tunnel (from dev box) diff --git a/tools/edge-control/install.sh b/tools/edge-control/install.sh index 68880ab..9571311 100755 --- a/tools/edge-control/install.sh +++ b/tools/edge-control/install.sh @@ -43,18 +43,21 @@ INSTALL_DIR="/opt/disinto-edge" REGISTRY_DIR="/var/lib/disinto" CADDY_VERSION="2.8.4" DOMAIN_SUFFIX="disinto.ai" +EXTRA_CADDYFILE="/etc/caddy/extra.d/*.caddy" usage() { cat <<EOF Usage: $0 [options] Options: - --gandi-token <token> Gandi API token for wildcard cert (required) - --install-dir <dir> Install directory (default: /opt/disinto-edge) - --registry-dir <dir> Registry directory (default: /var/lib/disinto) - --caddy-version <ver> Caddy version to install (default: ${CADDY_VERSION}) - --domain-suffix <suffix> Domain suffix for tunnels (default: disinto.ai) - -h, --help Show this help + --gandi-token <token> Gandi API token for wildcard cert (required) + --install-dir <dir> Install directory (default: /opt/disinto-edge) + --registry-dir <dir> Registry directory (default: /var/lib/disinto) + --caddy-version <ver> Caddy version to install (default: ${CADDY_VERSION}) + --domain-suffix <suffix> Domain suffix for tunnels (default: disinto.ai) + --extra-caddyfile <path> Import path for operator-owned Caddy config + (default: /etc/caddy/extra.d/*.caddy) + -h, --help Show this help Example: $0 --gandi-token YOUR_GANDI_API_TOKEN @@ -84,6 +87,10 @@ while [[ $# -gt 0 ]]; do DOMAIN_SUFFIX="$2" shift 2 ;; + --extra-caddyfile) + EXTRA_CADDYFILE="$2" + shift 2 + ;; -h|--help) usage ;; @@ -225,8 +232,29 @@ EOF chmod 600 "$GANDI_ENV" # Create Caddyfile with admin API and wildcard cert +# Note: Caddy auto-generates server names (srv0, srv1, …). lib/caddy.sh +# discovers the server name dynamically via _discover_server_name() so we +# don't need to name the server here. CADDYFILE="/etc/caddy/Caddyfile" -cat > "$CADDYFILE" <<EOF + +# Back up existing Caddyfile before overwriting +if [ -f "$CADDYFILE" ] && [ ! -f "${CADDYFILE}.pre-disinto" ]; then + cp "$CADDYFILE" "${CADDYFILE}.pre-disinto" + log_info "Backed up existing Caddyfile to ${CADDYFILE}.pre-disinto" +fi + +# Create extra.d directory for operator-owned site blocks +EXTRA_DIR="/etc/caddy/extra.d" +mkdir -p "$EXTRA_DIR" +chmod 0755 "$EXTRA_DIR" +if getent group caddy >/dev/null 2>&1; then + chown root:caddy "$EXTRA_DIR" +else + log_warn "Group 'caddy' does not exist; extra.d owned by root:root" +fi +log_info "Created ${EXTRA_DIR} for operator-owned Caddy config" + +cat > "$CADDYFILE" <<CADDYEOF # Caddy configuration for edge control plane # Admin API enabled on 127.0.0.1:2019 @@ -240,7 +268,10 @@ cat > "$CADDYFILE" <<EOF dns gandi {env.GANDI_API_KEY} } } -EOF + +# Operator-owned site blocks (apex, www, static content, etc.) +import ${EXTRA_CADDYFILE} +CADDYEOF # Start Caddy systemctl restart caddy 2>/dev/null || { @@ -359,6 +390,7 @@ echo "Configuration:" echo " Install directory: ${INSTALL_DIR}" echo " Registry: ${REGISTRY_FILE}" echo " Caddy admin API: http://127.0.0.1:2019" +echo " Operator site blocks: ${EXTRA_DIR}/ (import ${EXTRA_CADDYFILE})" echo "" echo "Users:" echo " disinto-register - SSH forced command (runs ${INSTALL_DIR}/register.sh)" diff --git a/tools/edge-control/lib/caddy.sh b/tools/edge-control/lib/caddy.sh index 69970cf..1e16cdc 100755 --- a/tools/edge-control/lib/caddy.sh +++ b/tools/edge-control/lib/caddy.sh @@ -19,6 +19,24 @@ CADDY_ADMIN_URL="${CADDY_ADMIN_URL:-http://127.0.0.1:2019}" # Domain suffix for projects DOMAIN_SUFFIX="${DOMAIN_SUFFIX:-disinto.ai}" +# Discover the Caddy server name that listens on :80/:443 +# Usage: _discover_server_name +_discover_server_name() { + local server_name + server_name=$(curl -sS "${CADDY_ADMIN_URL}/config/apps/http/servers" \ + | jq -r 'to_entries | map(select(.value.listen[]? | test(":(80|443)$"))) | .[0].key // empty') || { + echo "Error: could not query Caddy admin API for servers" >&2 + return 1 + } + + if [ -z "$server_name" ]; then + echo "Error: could not find a Caddy server listening on :80/:443" >&2 + return 1 + fi + + echo "$server_name" +} + # Add a route for a project # Usage: add_route <project> <port> add_route() { @@ -26,6 +44,9 @@ add_route() { local port="$2" local fqdn="${project}.${DOMAIN_SUFFIX}" + local server_name + server_name=$(_discover_server_name) || return 1 + # Build the route configuration (partial config) local route_config route_config=$(cat <<EOF @@ -58,16 +79,21 @@ add_route() { EOF ) - # Append route using POST /config/apps/http/servers/edge/routes - local response - response=$(curl -s -X POST \ - "${CADDY_ADMIN_URL}/config/apps/http/servers/edge/routes" \ + # Append route via admin API, checking HTTP status + local response status body + response=$(curl -sS -w '\n%{http_code}' -X POST \ + "${CADDY_ADMIN_URL}/config/apps/http/servers/${server_name}/routes" \ -H "Content-Type: application/json" \ - -d "$route_config" 2>&1) || { + -d "$route_config") || { echo "Error: failed to add route for ${fqdn}" >&2 - echo "Response: ${response}" >&2 return 1 } + status=$(echo "$response" | tail -n1) + body=$(echo "$response" | sed '$d') + if [ "$status" -ge 400 ]; then + echo "Error: Caddy admin API returned ${status}: ${body}" >&2 + return 1 + fi echo "Added route: ${fqdn} → 127.0.0.1:${port}" >&2 } @@ -78,31 +104,45 @@ remove_route() { local project="$1" local fqdn="${project}.${DOMAIN_SUFFIX}" - # First, get current routes - local routes_json - routes_json=$(curl -s "${CADDY_ADMIN_URL}/config/apps/http/servers/edge/routes" 2>&1) || { + local server_name + server_name=$(_discover_server_name) || return 1 + + # First, get current routes, checking HTTP status + local response status body + response=$(curl -sS -w '\n%{http_code}' \ + "${CADDY_ADMIN_URL}/config/apps/http/servers/${server_name}/routes") || { echo "Error: failed to get current routes" >&2 return 1 } + status=$(echo "$response" | tail -n1) + body=$(echo "$response" | sed '$d') + if [ "$status" -ge 400 ]; then + echo "Error: Caddy admin API returned ${status}: ${body}" >&2 + return 1 + fi # Find the route index that matches our fqdn using jq local route_index - route_index=$(echo "$routes_json" | jq -r "to_entries[] | select(.value.match[]?.host[]? == \"${fqdn}\") | .key" 2>/dev/null | head -1) + route_index=$(echo "$body" | jq -r "to_entries[] | select(.value.match[]?.host[]? == \"${fqdn}\") | .key" 2>/dev/null | head -1) if [ -z "$route_index" ] || [ "$route_index" = "null" ]; then echo "Warning: route for ${fqdn} not found" >&2 return 0 fi - # Delete the route at the found index - local response - response=$(curl -s -X DELETE \ - "${CADDY_ADMIN_URL}/config/apps/http/servers/edge/routes/${route_index}" \ - -H "Content-Type: application/json" 2>&1) || { + # Delete the route at the found index, checking HTTP status + response=$(curl -sS -w '\n%{http_code}' -X DELETE \ + "${CADDY_ADMIN_URL}/config/apps/http/servers/${server_name}/routes/${route_index}" \ + -H "Content-Type: application/json") || { echo "Error: failed to remove route for ${fqdn}" >&2 - echo "Response: ${response}" >&2 return 1 } + status=$(echo "$response" | tail -n1) + body=$(echo "$response" | sed '$d') + if [ "$status" -ge 400 ]; then + echo "Error: Caddy admin API returned ${status}: ${body}" >&2 + return 1 + fi echo "Removed route: ${fqdn}" >&2 } @@ -110,13 +150,18 @@ remove_route() { # Reload Caddy to apply configuration changes # Usage: reload_caddy reload_caddy() { - local response - response=$(curl -s -X POST \ - "${CADDY_ADMIN_URL}/reload" 2>&1) || { + local response status body + response=$(curl -sS -w '\n%{http_code}' -X POST \ + "${CADDY_ADMIN_URL}/reload") || { echo "Error: failed to reload Caddy" >&2 - echo "Response: ${response}" >&2 return 1 } + status=$(echo "$response" | tail -n1) + body=$(echo "$response" | sed '$d') + if [ "$status" -ge 400 ]; then + echo "Error: Caddy reload returned ${status}: ${body}" >&2 + return 1 + fi echo "Caddy reloaded" >&2 }