[DMC-010] Exit 2 is overloaded between the honest lane and argparse usage errors

[chaoz23]

Exit 2 currently carries two unrelated meanings:

1. the codified honest lane — unusable input/charter, a first-class cannot-adjudicate verdict that SKILL.md tells agents to route to a human without retrying
2. an ordinary argparse usage error — e.g. dmcheck --zzz-not-a-real-flag → exit 2

The two are distinguishable only by whether stdout happens to contain JSON.

An agent obeying SKILL.md escalates its own malformed invocation to a human as if it were a ruling — when the correct response is to fix the call and retry.

For contrast, srdcheck separates these cleanly: bad flags return 3, documented as "a usage/internal error (fix the call), never a verdict."

Suggested fix: move usage errors off exit 2 (srdcheck's exit-3 convention is the family precedent), or document the disambiguation rule explicitly in SKILL.md so agents can classify reliably.

Severity is deliberately low: the honest lane still works, and the JSON envelope disambiguates for anyone parsing stdout. Filed so it is not rediscovered.

---

Found by a conformance audit of the check-family four (srdcheck, charactercheck, dmcheck, table-kit) against FAMILY.md v1 on 2026-08-17. Audited at working-tree HEAD, which was 1–3 commits ahead of origin/main and 0 behind, so absences hold for main. Every claim below was verified by execution or by git grep across all local heads, then independently re-derived by a second pass instructed to refute it.

[chaoz23]

Reopened — this was closed in error by #17.

#17 corrected SKILL.md to name stdout as the envelope stream, and in doing so documented how to tell the two exit-2 lanes apart:

exit 2 with an envelope on stdout is the honest lane; exit 2 with empty stdout means the call was malformed

That makes the overload navigable. It does not remove it. Both conditions still return 2, and separating them is a code change — srdcheck's exit 3 for usage errors is the family precedent, per FAMILY.md clause 1 (see chaoz23/srdcheck#83, which states explicitly that usage errors must not share the honest lane's exit code).

Cause of the mis-close, recorded so it does not recur: #17's body contained the sentence "That does not close #15". GitHub's keyword parser matches close #15 and ignores the negation. Never write a closing keyword next to an issue reference you intend to leave open — say "#15 remains open" instead.