Your objective is to implement $meter waste [--window <N>] [--run-id <id>].
Before proceeding, verify that .cocoplus/ exists. If not, output: "CocoPlus not initialized in this directory. Run $pod init to begin." Then stop.
Contract
Read recent session, meter, and reconciliation artifacts. Estimate avoidable context reconstruction, repeated failed loops, off-brief execution, and stale-context reloads. Use transcript-derived reconciliation when present; otherwise label the result as estimate-only.
Output:
- Waste window and source files
- Waste categories and approximate token share
- Highest-confidence cause
- Suggested prevention
- Candidate wisdom or style rule, if evidence is strong
Do not write directly to do-not-use.md. If a rejected approach seems warranted, append a review candidate to .cocoplus/wisdom/review-queue.md so the operator can accept it with $wisdom reject.
Exit Criteria
- The report separates verified waste from estimated waste.
- Off-brief and loop waste are identified when evidence exists.
- Recommendations prefer correctness before token savings.
- Negative-memory candidates remain review-only until explicitly accepted.
Anti-Rationalization
| Shortcut / Temptation | Why It Fails |
|---|---|
| Treat the skill as complete because the file exists | Skill contracts must describe observable behavior and verification, not just command names. |
| Skip artifact and safety checks for a small command | Small commands still mutate state or guide execution; preserve the same gates. |