SPEC-001 fails its own validator: a document about writing specs, filed as a spec

[explosivebit]

Found while implementing #450 (PR #472, merged as 551ddb9). Deliberately not fixed there.

What

forgeplan validate SPEC-001 fails two rules:

x [MUST] spec-summary
x [MUST] spec-contracts

It failed before #472 and still fails after. The other five specs (SPEC-002 … SPEC-006) pass.

Why this is not a validator bug

SPEC-001 is a document about how to write specifications. It has no API contract and no behavioural contract because it does not specify a system — it explains a form. The validator is right to say a spec must contain a contract; SPEC-001 is right to have none. The mismatch is the filing, not the rule.

This was checked before loosening anything. #472 widened spec-contracts to accept behavioural contracts (## Requirements / ## Contract / ## Behavioral Contract) alongside structural ones — SPEC-001 satisfies none of the five, and adding a sixth escape hatch to accommodate one meta-document would be the wrong trade.

Options

1. Re-file as a Note or a doc under docs/methodology/ and supersede SPEC-001 with a pointer. Cleanest: the content is guidance, and guidance already has a home.
2. Leave it and accept a permanently red artifact. Costs nothing today, but a validator finding that is understood-and-ignored trains everyone to ignore validator findings.

Option 1 is recommended. Whoever takes it should use supersede, not delete — the content is still useful, only misfiled.

Not urgent

Nothing is blocked by it. Filed because "SPEC-001 fails validation" will otherwise be rediscovered by the next person who runs validate --all and re-derives this reasoning from scratch.

Refs: #450, #472, prob-105