mece-logic-checkerlisted

# MECE Logic Checker (WiseTalk Skill-4) ## Important - The verdict comes **only** from `scripts/mece-check.py`'s deterministic output. Never invent an overlap reason or a missing dimension — if the script fails, report the failure instead. - Never claim a list is MECE (`is_valid: true`) without having run the script this run. - The fixed dimension library is the 4M1E set: **Human, Machine, Material, Method, Environment** — it is the only omission standard (D-003). ## Instructions ### Step 1: Extract the argument points Identify each distinct argument point in the user's message. If the user pasted prose, split it into one point per argument. Drop empty entries. Done when every point is a single clean string and the list has no empties. If fewer than 2 non-empty points remain, stop and report `{"message": "Cannot perform MECE analysis"}` — the script will also enforce this. ### Step 2: Run the checker Pass the points as a JSON array: python scripts/mece-check.py --points '["Point A", "Point B", "Point C"]' Expected output: a single JSON line — `{"is_valid": bool, "overlap_reason": str|null, "missing_dimension": str|null}`, or `{"message": "Cannot perform MECE analysis"}` for fewer than 2 points. If it fails: script not found → check the path `scripts/mece-check.py` relative to this skill folder; JSON quoting error (exit 2) → pass the array as valid JSON with single-quoted wrapping and double-quoted strings; Python unavailable → go to Fallback. ### Step 3: Interpret