fix: feat(20g): migrate all remaining agents to .profile + remove ops repo journal dirs (#90)

AGENTS.md

@@ -39,9 +39,6 @@ disinto-ops/             (ops repo — {project}-ops)
 │   ├── approved/  approved vault items
 │   ├── fired/     executed vault items
 │   └── rejected/  rejected vault items
-├── journal/
-│   ├── planner/   daily planning logs
-│   └── supervisor/ operational health logs
 ├── knowledge/     shared agent knowledge + best practices
 ├── evidence/      engagement data, experiment results
 ├── portfolio.md   addressables + observables
@@ -49,6 +46,45 @@ disinto-ops/             (ops repo — {project}-ops)
 └── RESOURCES.md   accounts, tokens (refs), infra inventory
 ```
 
+> **Note:** Journal directories (`journal/planner/` and `journal/supervisor/`) have been removed from the ops repo. Agent journals are now stored in each agent's `.profile` repo on Forgejo.
+
+## Agent .profile repos
+
+Each agent maintains a `.profile` repo on Forgejo that stores:
+- `formula.toml` — Agent-specific formula (overrides local `formulas/` defaults)
+- `knowledge/lessons-learned.md` — Aggregated lessons from journal digests
+- `journal/` — Per-session reflection journals (archived after digestion)
+
+### Formula resolution
+
+Agents load their formula from `.profile` first, falling back to local `formulas/`:
+
+```bash
+load_formula_or_profile "agent-role" "formulas/agent-role.toml"
+```
+
+### Lessons injection
+
+At session start, agents load `knowledge/lessons-learned.md` from `.profile` and inject it into the prompt:
+
+```bash
+formula_prepare_profile_context
+```
+
+This single function call replaces the previous boilerplate of `profile_load_lessons` + `LESSONS_INJECTION` assignments.
+
+### Journal writing
+
+After each session, agents write reflection journals to `.profile/journal/`:
+
+```bash
+profile_write_journal "session-name" "Session title" "outcome" "files-changed"
+```
+
+Journals are automatically digested into `lessons-learned.md` when undigested count exceeds 10.
+
+See [lib/formula-session.sh](lib/formula-session.sh) for the full `.profile` API reference.
+
 > **Terminology note:** "Formulas" in this repo are TOML issue templates in `formulas/` that
 > orchestrate multi-step agent tasks (e.g., `run-gardener.toml`, `run-planner.toml`). This is
 > distinct from "processes" described in `docs/EVIDENCE-ARCHITECTURE.md`, which are measurement

bin/disinto

@@ -896,8 +896,6 @@ setup_ops_repo() {
   mkdir -p "${ops_root}/vault/approved"
   mkdir -p "${ops_root}/vault/fired"
   mkdir -p "${ops_root}/vault/rejected"
-  mkdir -p "${ops_root}/journal/planner"
-  mkdir -p "${ops_root}/journal/supervisor"
   mkdir -p "${ops_root}/knowledge"
   mkdir -p "${ops_root}/evidence/engagement"
 
@@ -916,9 +914,6 @@ ${ops_name}/
 │   ├── approved/         # approved vault items
 │   ├── fired/            # executed vault items
 │   └── rejected/         # rejected vault items
-├── journal/
-│   ├── planner/          # daily planning logs
-│   └── supervisor/       # operational health logs
 ├── knowledge/            # shared agent knowledge and best practices
 ├── evidence/             # engagement data, experiment results
 ├── portfolio.md          # addressables + observables