grill

Pre-capture interrogation. Pressure-tests a plan or decision before it's frozen into an artifact, then routes to the right capture skill. Writes nothing, touches nothing under .task/.

See the grill guide for the full walkthrough.

Usage

/task:grill [<context>]

Input — $ARGUMENTS, optional. A topic or free-form context to grill ("the retry design", "whether to shard the queue"). Empty → grills the plan being discussed in the current chat.

What it does

• Frames the target in a sentence or two.
• Resolves facts itself (from the codebase) and spends questions only on genuine decisions.
• Grills one AskUserQuestion fork at a time — each with 2–4 concrete options and an honest recommendation.
• Keeps a decision ledger: {the decision} — because {the reason}.
• Ends with a pre-mortem kill-shot question — skipped when no answer to it would change the ledger — then prints the full ledger.

Depth is an outcome, not a target — it stops when no unanswered fork would change the capture, and offers a wrap-up-or-keep-going choice when only secondary forks remain.

Config

No config gate, no setup. grill is the one skill that neither reads nor writes anything under .task/ — it runs in a fresh, unconfigured project. Dialog mirrors the language of your chat.

Output

A decision ledger printed as chat text (there is no file), followed by a routing footer naming exactly one next skill:

## Decision ledger
1. {decision at full specificity} — because {the load-bearing reason}
2. …

→ Next: /task:to-plan — one task with the approach nailed down; capture Description + Plan.

Does not

• Write any file, anywhere — serializing the ledger is the to-* skills' job.
• Batch questions — one fork per prompt, always.
• Implement, refactor, or "just fix it" mid-grill.
• Agree by reflex — if your leaning is the weaker call, the recommendation says so.