## Summary
- Claude Code v2.1.79 permanently shows `❯` in the input area even while actively thinking, causing `monitor_phase_loop` to false-positive on idle detection and kill working sessions after 90 seconds
- Replace `tmux capture-pane | grep ❯` with a Claude Code Stop hook (`lib/hooks/on-idle-stop.sh`) that writes a marker file only when Claude actually finishes responding
- Hook is installed per-worktree in `.claude/settings.json` by `create_agent_session`; marker cleaned up on inject/kill
## Test plan
- [x] Verified hook installs correctly in fresh worktree
- [x] Verified marker file appears only after Claude finishes responding (not during active thinking)
- [x] Verified live dev-agent session picks up fix and Claude works without being killed
- [x] Verified `agent_inject_into_session` clears marker before new work
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: openhands <openhands@all-hands.dev>
Reviewed-on: https://codeberg.org/johba/disinto/pulls/272
Add diff_has_code_files() and ci_required_for_pr() helpers to
ci-helpers.sh. Non-code PRs (docs/*, formulas/*, evidence/*, *.md)
that have no CI results now skip the CI gate instead of being stuck
forever.
Applied to:
- review-pr.sh: CI gate skipped for non-code PRs
- review-poll.sh: CI gate skipped for non-code PRs
- dev-poll.sh: CI state treated as "success" for non-code PRs in
orphan, stuck-PR, and backlog merge paths
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Fixes#261
## Changes
Fixed gardener hanging forever when Claude skips phase protocol. Three changes: (1) gardener-agent.sh: replaced 999999s timeout with 7200s (2h, matching dev-agent); (2) lib/agent-session.sh: added idle-prompt detection to monitor_phase_loop — if Claude returns to the ❯ prompt for 3 consecutive polls with no phase file written, exits immediately with _MONITOR_LOOP_EXIT=idle_prompt (only fires when phase file is empty, so awaiting_ci/review waits are unaffected); (3) gardener prompt: removed 'no time limit' wording, replaced with explicit phase-write requirement.
Co-authored-by: openhands <openhands@all-hands.dev>
Reviewed-on: https://codeberg.org/johba/disinto/pulls/263
Reviewed-by: Disinto_bot <disinto_bot@noreply.codeberg.org>
- supervisor-poll.sh: check PR state before retrigger; discard stale escalations
for closed/merged PRs instead of pushing to their branches
- supervisor-poll.sh: bump escalation ts to now on failed retrigger push, so
the 30-min cooldown resets and alert flooding is avoided on persistent failures
- ci-helpers.sh: require at least one confirmed infra step before returning
"infra"; prevents false-positive when all step names are empty strings
- ci-helpers.sh: clarify header comment to distinguish per-function requirements
- AGENTS.md: document classify_pipeline_failure() in ci-helpers.sh table row
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add optional second TOKEN parameter to codeberg_api_all in lib/env.sh,
defaulting to $CODEBERG_TOKEN. Pass $REVIEW_BOT_TOKEN at the dismiss
block in review-pr.sh so reviews are fetched as the review bot account.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add optional 4th parameter to monitor_phase_loop for SESSION_NAME,
falling back to the $SESSION_NAME global for backwards-compatibility.
Document the full function signature in both the file header and inline comment.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Export LAST_PHASE_MTIME from monitor_phase_loop before invoking the callback
so that phase-handler.sh can compare phase file mtimes inside the awaiting_review
inner poll loop without hitting an unbound variable error under set -u.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Library functions need explicit session name argument — they no longer
have closure over $SESSION_NAME from the parent script.
- agent_kill_session: add $SESSION_NAME to all 11 call sites
- agent_inject_into_session: add $SESSION_NAME to all call sites in
phase-handler.sh and gardener-agent.sh
- agent_kill_session: guard against missing arg (defensive)
Both dev-agent.sh and gardener-agent.sh call these functions but they
were never implemented during the #158 extraction. Adds:
- create_agent_session(session, workdir) — tmux + claude + wait for ready
- inject_formula(session, text) — alias for agent_inject_into_session
Remove || break from the codeberg_api call in the pagination loop.
With set -euo pipefail in all callers, a failed fetch now exits the
function non-zero — matching the original curl -sf behaviour where a
network or auth error aborted the script rather than returning empty
results and risking a duplicate review.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Fix SC2164: add || exit 1 to bare cd in update-prompt.sh
- Fix SC2155: separate declare and assign in env.sh, supervisor-poll.sh, dev-agent.sh
- Fix SC2034: inline suppression for vars used by sourced helpers
- Remove unused `mergeable` declaration, rename unused loop var to `_w`
- Remove || true from shellcheck CI step — failures are now blocking
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Fix double-injection bug: flat-file write only when direct tmux inject didn't happen
- Fix ci_exhausted href='#' fallback to use CODEBERG_WEB/pulls/N
- Remove duplicate $THREAD_FILE in rm command
- HTML-escape CI snippet before embedding in <pre> block
- notify_ctx falls back to plain matrix_send when no thread exists
- Thread root uses HTML-formatted message for consistency
- Deduplicate _ci_pipeline_url variable
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Guard inject_into_session wait_for_claude_ready with || true
- Guard all tmux calls in inject_into_session with || true
- Add worktree cleanup to idle-timeout branch in review-poll.sh
- Check phase before sleep in wait_for_review_output (no 10s delay)
- Prune review-thread-map entries during session cleanup
- Skip human question injection during active review (phase check)
- Remove no-op tmux kill-session after has-session returns false
- Add ASCII fallback for Claude prompt detection (locale safety)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Rewrite review-pr.sh to use persistent tmux sessions instead of one-shot
claude -p. Re-reviews inject incremental diffs into the same session so
Claude remembers what it flagged and can verify fixes were addressed.
- review-pr.sh: tmux session review-{project}-{pr}, phase protocol
(PHASE:review_complete), JSON output via file, retry on invalid JSON
- review-poll.sh: session lifecycle cleanup (merged/closed PRs, 4h idle)
- matrix_listener.sh: route human questions to review sessions via
/tmp/review-thread-map
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Move atomic mv inside gardener loop so reply is only claimed when a
matching needs_human session exists (fixes reply-loss regression)
- Delay rm of claimed file until after successful injection in both
supervisor and gardener (OOM/SIGKILL leaves file recoverable)
- Fix matrix_listener ack message: 'next poll' instead of 'next supervisor poll'
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Implements the vault subsystem: a JSONL queue and gate agent that sits
between agent output and irreversible external actions (emails, posts,
API calls, charges).
New files:
- vault/vault-poll.sh: cron entry (*/30), three phases: retry approved,
timeout escalations (48h), invoke vault-agent for new pending actions
- vault/vault-agent.sh: claude -p wrapper that classifies and routes
actions based on risk × reversibility routing table
- vault/vault-fire.sh: two-phase dispatcher (pending→approved→fired)
with per-action locking and webhook-call handler
- vault/vault-reject.sh: moves actions to rejected/ with reason + timestamp
- vault/PROMPT.md: vault-agent system prompt with routing table
Modified:
- lib/matrix_listener.sh: new vault dispatch branch for APPROVE/REJECT
replies to escalation threads
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace lib/parse-deps.py with lib/parse-deps.sh to keep the toolchain
all-bash. Rewrite supervisor P3b cycle detection and P3c stale dep check
as pure bash using associative arrays and DFS.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Single source of truth for dependency parsing, replacing three copies:
- dev-poll.sh get_deps() now calls parse-deps.py
- supervisor P3b/P3c import parse_deps() via importlib
Supports stdin, argument, and --json modes for different callers.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The supervisor agent was confusingly named "factory" (same as the
project). Rename directory, script, log, lock, status, and escalation
files. Update all references across scripts and docs.
FACTORY_ROOT env var unchanged (refers to project root, not agent).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Dev-agent failed with exit 127 (command not found) because claude
is installed in ~/.local/bin which wasn't in the PATH set by env.sh.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add matrix_send() to lib/env.sh and matrix_listener.sh daemon for
real-time notifications, threaded escalations, and human-in-the-loop
replies. All agents now notify via Matrix instead of openclaw.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Moved ci-debug.sh from dev/ to lib/ (shared utility)
- README: fixed supervisor description (all alerts go to claude)
- README: replaced implementation details with actual design principles
