| name | plan |
| description | Analyze the full ecosystem and surface what needs doing next — including emergent tasks not on the board |
| user-invocable | true |
| disable-model-invocation | true |
| allowed-tools | Bash, Read, Grep, Glob, Agent |
Analyze the ecosystem and produce an actionable plan. This is a read-only operation — do NOT fix anything, do NOT add board items, do NOT modify files.
Arguments
- No argument: Run all 5 analysis layers
conventions: Run only the convention audit layer with full detail
Analysis Layers
If argument is conventions:
Spawn the convention-auditor agent and wait for its result. Display the full convention audit table. Done.
If no argument (full analysis):
Run these 5 layers, parallelizing where possible:
Layer 1 — Board State (parallel with 2 and 3)
Fetch the board:
"/c/Program Files/GitHub CLI/gh.exe" project item-list 1 --owner pon00050 --format json
Check for:
- Items stuck In Progress for >7 days (stale)
- Items recently moved to Todo that were previously Blocked
- Total counts by status and owner
Layer 2 — Repo Health (parallel with 1 and 3)
For each repo in C:\Users\pon00\Projects\ (krff-shell, kr-forensic-core, kr-dart-pipeline, kr-anomaly-scoring, kr-stat-tests, kr-company-registry, kr-trading-calendar, kr-beneish, kr-derivatives, jfia-catalog, jfia-forensic, kr-enforcement-cases, kr-real-estate):
git -C ../REPO status --porcelain
git -C ../REPO log --oneline -1 --format="%cr"
git -C ../REPO rev-list --count @{u}..HEAD 2>/dev/null
Flag:
- Uncommitted changes
- Unpushed commits
- Stale repos (no commits in 14+ days)
- Known Gaps with "bug" or "risk" in status suffix (code quality signal)
- Repos missing a
## Known Gaps section in CLAUDE.md (convention drift)
Layer 3 — Convention Drift (parallel with 1 and 2, via background agent)
Spawn the convention-auditor agent in the background:
Use Agent tool with subagent_type convention-auditor, run_in_background: true
Continue with layers 1-2 while this runs. Incorporate results at the end.
Layer 4 — Integration Gaps (after layers 1-2)
Check across repos:
- Repos with
tests/ but no conftest.py
- Repos with
src/ but no _paths.py or paths.py
- Repos where CLAUDE.md references a test command that doesn't match the actual test structure
- kr-derivatives data path consistency with krff-shell/kr-dart-pipeline outputs
Use Glob and Grep across C:\Users\pon00\Projects\*.
Layer 5 — Strategic Alignment (last, synthesizes everything)
Read ECOSYSTEM.md in the hub. Cross-reference with board state from Layer 1:
- Are there ECOSYSTEM.md backlog items not on the board?
- Are there board items that are newly unblocked (dependencies completed)?
- What P1/P2 items could be started now?
Output Format
=== Ecosystem Plan ===
Board: X todo (Y AI, Z human), W done
Health: X/8 repos clean, Y need attention
Emergent Tasks (not on board):
1. [HIGH] description — repo — rationale
2. [MED] description — repo — rationale
...
Suggested Next Session:
AI: [specific task] — [why now]
Human: [specific task] — [deadline if any]
Convention Drift: X deviations found (Y since last audit)
[summary table from convention-auditor]
Rules
- Report only. Do NOT fix anything, do NOT add board items, do NOT modify any files.
- Be specific. Emergent tasks must include the repo, the file(s) affected, and why it matters.
- Prioritize actionability. Don't list 20 minor things. Surface the 3-5 most impactful items.
- Convention audit details go in the convention drift section. Keep the main output focused on emergent tasks and next-session suggestions.