agentleFS
Sign inStart free
IgorGanapolskyIgorGanapolsky/ThumbGate/.agents/skills/craft-playbook-gate/SKILL.md

craft-playbook-gate

Skill29 starsChanged 7 months agoMITView on GitHub

Rachel Woods's autonomous-AI episode is a playbook format, not a ThumbGate product. Keep one atomic playbook in git, refuse zero-touch on the first cycle, and require a backup hash plus a named approval before a structural self-edit. Use when checking a playbook outline, a CRAFT cycle, a self-edit, a trigger word, or a local stack. Do NOT use to install dayglass or skill-up, store the playbook in a vendor chat UI, spawn a subagent, call Asana or ClickUp, reply on YouTube, or email the host.

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep one deliverable's trigger, inputs, steps, output, and escalation in git.

    line 17

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Craft playbook gate — own the playbook, rent the model

Goal

Keep one deliverable's trigger, inputs, steps, output, and escalation in git. The first cycle keeps a named human step. A later cycle can close that step only after every CRAFT stage is done and each locked win is text inside a step.

Constraints

NEVERALWAYS
Install dayglass or skill-upnode scripts/craft-playbook-gate.js
Use Docker or OpenSandboxHome git and a relative .md or .json path
Store the playbook in a vendor chat UIBoolean stages, in order: clear, realistic, setup, feedback, rollout
Approve a self-edit by line count or percentA backup sha256 of the previous playbook
Spawn a subagent or call Asana or ClickUpExplicit approved: true and a named approvedBy for a structural edit
Claim zero-touch on cycle 1Trigger word nitpicking, not feedback
Reply on the YouTube episode or email the hostA stack is done only when every child has verified: true

Procedures

node scripts/craft-playbook-gate.js --map-only
node scripts/craft-playbook-gate.js
npm run craft:playbook
npm run test:craft-playbook-gate

--help and --map-only print and exit without reading stdin. Any other command reads one JSON object from stdin. The command writes nothing. Exit 0 is accept. Exit 1 is a hold. A minor detail tweak may be applied later only when the backup hash matches and feedbackRef names the review. Changing the trigger, a step id, an escalation flag, or the escalation text stays held until a person approves it. Moving the playbook out of git, or onto an absolute path, is a hold even when the rest of the edit is minor.

Code as truth: scripts/craft-playbook-gate.js Provenance: https://www.youtube.com/watch?v=CMkwClAm2TI

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free