Clarify scheduler log-directory bootstrap and first-run behavior

src/prompts/META_PROMPTS.md

@@ -25,14 +25,15 @@ Use the `locked` array from the JSON output as the exclusion set.
 
 MUST 3: Run these commands in this order:
 1) test -f AGENTS.md && cat AGENTS.md || echo "No AGENTS.md found; continuing" (missing AGENTS.md is non-fatal)
-2) ls -1 task-logs/daily/ | sort | tail -n 1
+2) mkdir -p task-logs/daily task-logs/weekly
-3) Select next roster agent not in exclusion set
+3) ls -1 task-logs/daily/ | sort | tail -n 1
-4) Claim via repository lock:
+4) Select next roster agent not in exclusion set (empty directory means first run; choose first eligible roster entry)
+5) Claim via repository lock:
    AGENT_PLATFORM=<platform> npm run lock:lock -- --agent <agent-name> --cadence daily
-   Exit 0 = lock acquired, proceed. Exit 3 = race lost, go back to step 2.
+   Exit 0 = lock acquired, proceed. Exit 3 = race lost, go back to step 3.
-5) Execute selected prompt from src/prompts/daily/
+6) Execute selected prompt from src/prompts/daily/
-6) Run repository checks (for example: npm run lint)
+7) Run repository checks (for example: npm run lint)
-7) Create `_completed.md` or `_failed.md`, commit, push
+8) Create `_completed.md` or `_failed.md`, commit, push
 
 MUST 4: If all daily agents are excluded, stop and write `_failed.md` with this exact reason: `All roster tasks currently claimed by other agents`.
 ```
@@ -58,14 +59,15 @@ Use the `locked` array from the JSON output as the exclusion set.
 
 MUST 3: Run these commands in this order:
 1) test -f AGENTS.md && cat AGENTS.md || echo "No AGENTS.md found; continuing" (missing AGENTS.md is non-fatal)
-2) ls -1 task-logs/weekly/ | sort | tail -n 1
+2) mkdir -p task-logs/daily task-logs/weekly
-3) Select next roster agent not in exclusion set
+3) ls -1 task-logs/weekly/ | sort | tail -n 1
-4) Claim via repository lock:
+4) Select next roster agent not in exclusion set (empty directory means first run; choose first eligible roster entry)
+5) Claim via repository lock:
    AGENT_PLATFORM=<platform> npm run lock:lock -- --agent <agent-name> --cadence weekly
-   Exit 0 = lock acquired, proceed. Exit 3 = race lost, go back to step 2.
+   Exit 0 = lock acquired, proceed. Exit 3 = race lost, go back to step 3.
-5) Execute selected prompt from src/prompts/weekly/
+6) Execute selected prompt from src/prompts/weekly/
-6) Run repository checks (for example: npm run lint)
+7) Run repository checks (for example: npm run lint)
-7) Create `_completed.md` or `_failed.md`, commit, push
+8) Create `_completed.md` or `_failed.md`, commit, push
 
 MUST 4: If all weekly agents are excluded, stop and write `_failed.md` with this exact reason: `All roster tasks currently claimed by other agents`.
 ```

src/prompts/scheduler-flow.md

@@ -22,16 +22,25 @@ Use this document for all scheduler runs.
    test -f AGENTS.md && cat AGENTS.md || echo "No AGENTS.md found; continuing"
    ```
 
-4. Find latest cadence log file, then choose next roster agent not in exclusion set:
+4. Bootstrap log directories before listing files:
+
+   ```bash
+   mkdir -p <log_dir>
+   ```
+
+5. Find latest cadence log file, then choose next roster agent not in exclusion set:
 
    ```bash
    ls -1 <log_dir> | sort | tail -n 1
    ```
 
-5. If every roster agent is excluded, write a `_failed.md` log with:
+   - An empty `<log_dir>` is valid and means there are no prior cadence logs yet (first run).
+   - On first run, select the first eligible roster entry that is not in the exclusion set.
+
+6. If every roster agent is excluded, write a `_failed.md` log with:
    `All roster tasks currently claimed by other agents` and stop.
 
-6. Claim selected agent:
+7. Claim selected agent:
 
    ```bash
    AGENT_PLATFORM=<platform> \
@@ -42,6 +51,6 @@ Use this document for all scheduler runs.
    - Exit `3`: race lost/already locked, return to step 2.
    - Exit `2`: lock backend error, write `_failed.md` with reason `Lock backend error`, stop.
 
-7. Execute `<prompt_dir>/<prompt-file>` end-to-end.
+8. Execute `<prompt_dir>/<prompt-file>` end-to-end.
 
-8. Create exactly one final status file (`_completed.md` or `_failed.md`), run repo checks, commit, and push.
+9. Create exactly one final status file (`_completed.md` or `_failed.md`), run repo checks, commit, and push.