Draft an approach, attack it with fresh eyes, and ask you before any work starts — the alignment gate.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Grill — the alignment gate
Before any work starts: draft an approach, **attack it**, then ask you.
This is the cheapest correction in the whole loop. A plan is a paragraph; the
work is an afternoon. Finding the missing file, the unhandled case or the absent
test while it is still a paragraph costs a few seconds of model time. Finding it
after the work is done costs the work.
## What it does
1. **Asks what you want planned.** Nothing runs until you have said.
2. **Drafts an approach** — concrete: which files, in what order, and the one
test that would prove it worked.
3. **Grills the draft.** A second pass whose only job is to find what the plan
fails to name. It did not write the plan and does not defend it.
4. **Checks the grill itself.** A critique that objects in general terms is
not usable — it cannot be acted on and cannot be argued with. This check
runs on a fresh context that has seen neither the plan-writer nor the
critic, and blocks if the grill did not quote what it was attacking.
5. **Asks you** before anything proceeds.
## Why the grill is a separate step
Every serious result on AI self-review says the same thing: a model reviewing its
…
The line at the top of this page pins the label v1.0.0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
GET https://markdownregistry.com/api/v1/artifacts/art_5adzoepe54tvxtuq
GET https://markdownregistry.com/api/v1/resolve?ref=vodouai/os/grill
GET https://markdownregistry.com/api/v1/blob/a9fbc352f797e68146460337151d4748673ef25049eb6d19cb5b0512f7c0d272
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
vodouai/os · integrations/claude-code-plugin/skills/vodou-memory/SKILL.md · Use the user's Vodou memory — recall what they told any AI before, and save what matters now. Use when the user…
vodouai/os · integrations/claude-code-plugin/skills/vodou-setup/SKILL.md · Install or connect the Vodou engine so this plugin's memory tools and hooks work. Use when the session-start hook…
vodouai/os · skills/agents/design/visual-storyteller/SKILL.md · Visual storytelling agent that structures presentations, plans data visualizations, builds case study narratives, and…
vodouai/os · skills/agents/engineering/ai-engineer/SKILL.md · Expert AI engineer for LLM integration, RAG pipelines, prompt engineering, model evaluation, and AI agent development
sethgammon/citadel · skills/grill/SKILL.md · Relentless one-question-at-a-time interview that sharpens a vague plan or design into shared understanding before any…
kerberosclaw/kc_ai_skills · grill/SKILL.md · Use when the user says 先討論 / wants to align on a fuzzy idea before building, or before any non-trivial implementation…
kanevry/session-orchestrator · skills/grill/SKILL.md · Use when the user wants a plan, design, or PRD stress-tested before any build — relentlessly interrogates one decision…
kanevry/session-orchestrator · .cursor/skills/grill/SKILL.md · Use when the user wants a plan, design, or PRD stress-tested before any build — relentlessly interrogates one decision…
kanevry/session-orchestrator · .codex-plugin/skills/grill/SKILL.md · Use when the user wants a plan, design, or PRD stress-tested before any build — relentlessly interrogates one decision…
kanevry/session-orchestrator · .agents/skills/grill/SKILL.md · Use when the user wants a plan, design, or PRD stress-tested before any build — relentlessly interrogates one decision…
technicalpickles/pickled-claude-plugins · plugins/stay-principled/skills/grill/SKILL.md · Walk a depth-first decision tree anchored to a project's documented principles before brainstorming opens options…
spair/task-pipeline · skills/grill/SKILL.md · Interrogate a plan or decision one question at a time before capture, keeping a decision-plus-rationale ledger, then…
