/hard-cheese

Use this skill when the human author needs to demonstrate they understand the causal logic of a diff before it is shared for review or a pull request is opened. The gate exists to mitigate epistemic debt — the failure mode where AI-scaffolded code passes review, type-checks, and tests green while the author cannot explain it to a reviewer.

Do not use it for code review (/age covers eight orthogonal review dimensions), for test hardening (/press), or for fix application (/cure). Those skills review the artifact. Hard-cheese reviews the human's understanding of the artifact — orthogonal axis.

Inputs

/hard-cheese [<slug>] [--socratic-cap N=3] [--no-judge]

Arguments:

• <slug> — optional. Identifies the artifact at .cheese/hard/<slug>.md. When omitted, fall back to the git short SHA of HEAD. An explicit slug always wins.
• --socratic-cap N — max retry attempts before the gate marks the artifact FAILED and exits non-zero. Default 3. Vibecheck does not cap; easy-cheese does to avoid infinite loops.
• --no-judge — log-only mode. Capture the user's explanation, write the artifact with status: LOGGED, skip the judge sub-agent spawn. Mirrors vibecheck's optional JSONL telemetry mode.

Invocation modes