agentsclimarketplace

Aa grill

Skill Peeyushmeher/agent-agile/skills/aa-grill

Agile planning + execution for AI coding agents: grill the idea, plan with OKRs, adversarial critic panel with kill authority, cheap-model story cards, parallel wave execution. MIT, no token, ever.

Install
npx -y skills add Peeyushmeher/agent-agile --skill aa-grill

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 24 days oldThe repository was created 24 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when the user has a rough or vague product idea that needs interrogation before any planning — or invokes /aa-grill. Locks the idea into .planning/IDEA.md or recommends against building it.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.5 KB, as published. Nobody here has run it

aa-grill

Resolve the Agent-Agile playbook root: use the first of these that exists — (1) ${CLAUDE_PLUGIN_ROOT}/playbooks, (2) ./.claude/agent-agile/playbooks, (3) ./.agents/agent-agile/playbooks, (4) ~/.claude/agent-agile/playbooks, (5) ~/.agents/agent-agile/playbooks, (6) ./playbooks.

Read playbooks/planner.md section The Grill in full, including the Grill → intake mapping table inside it, and follow it exactly — do NOT re-derive or improvise the system. The interrogation script, its handling rules, and the valid "recommend not building it" outcome are all defined there; use them as written.

How this skill runs

This is a live conversation, not a subagent job — it runs in the main loop. Ask the seven grill questions from the playbook one at a time, in order. Never dump the list at once. Apply the playbook's handling rules on every answer: a vague answer gets narrowed into a forced choice built from the answer's own words before moving on; "everyone" and "it'd be cool" are rejected outright per the playbook.

If the idea fails its own questions — no real user, no real switch reason, no number that would move — say so plainly, name which questions it failed, and recommend against building it. Stop there; do not write IDEA.md for an idea you've just recommended against.

Closing the grill

Once all seven questions produce real answers:

  1. Write .planning/IDEA.md from the template at playbooks/templates/IDEA.md, filling Problem, Exact user, The one action, Why now / why switch, Non-goals, and Success signal from the grill answers. Once written, IDEA.md is locked — do not reinterpret it later.
  2. Update .planning/STATE.md from the template at playbooks/templates/STATE.md: set Project to a short working name for the idea (confirm it with the user in one breath before writing it), Phase to grill, Now to "grill complete, IDEA.md locked", Next to "run /aa-new-project", Pointers to .planning/IDEA.md, and Blockers to none.
  3. Tell the user the grill is done and offer /aa-new-project as the next step. State, using the Grill → intake mapping table, exactly which of the 8 intake questions are already pre-filled from this conversation and which three (hard constraints, riskiest unknown, review-batch appetite) will still be asked fresh.

Gives 0 of the 12 instructions most review quality skills give

Counted across 1,048 of the 1,783 authors here whose files we hold, read 2026-08-06

  • ask questions one at a timein 82 of 1048, across 54 files
  • provide a recommended answer for each questionin 73 of 1048, across 45 files
  • explore the codebase instead of asking answerable questionsin 66 of 1048, across 37 files
  • resolve dependencies between decisions one-by-onein 42 of 1048, across 15 files
  • interview the user relentlessly about the planin 39 of 1048, across 12 files
  • order findings by severityin 29 of 1048
  • resolve each branch of the decision treein 28 of 1048, across 5 files
  • run a grilling sessionin 26 of 1048, across 5 files
  • update CONTEXT.md immediately when a term is resolvedin 26 of 1048, across 9 files
  • propose precise canonical terms for vague languagein 25 of 1048, across 6 files
  • create documentation files lazilyin 24 of 1048, across 5 files
  • use the domain-modeling skillin 22 of 1048, across 3 files

Said here and by no other author read

  • resolve the playbook root path
  • read the planner grill section in full
  • follow the playbook grill instructions exactly
  • apply the playbook handling rules to every answer
  • force a choice for vague answers
  • recommend against building failed ideas

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.