Specifying User-Thrown Gates

When this skill runs

Exactly one of:

1. A user-gate task has "requiresUserSpecification": true in its json:metadata fence, OR
2. The agent ran the "do I know HOW?" self-check (see executing-plans) and concluded the verification mechanics are ambiguous, OR
3. The user ran /specify-gate <task-id> manually.

In all other cases — where verifyCommand is concrete and every acceptanceCriteria has an observable proof — the agent executes the gate directly and does NOT invoke this skill.

Announce at start: "I'm using the specifying-gates skill to lock down verification mechanics for Task N."

CRITICAL — what this skill does NOT do

• Does not run the verification command. Specification only.
• Does not close the task. It only enriches metadata; the agent returns to executing-plans afterward.
• Does not re-run brainstorming. The design is already decided; only the HOW of one gate is missing.