agentleFS
rossinsilicorossinsilico/loop-anything/.agents/skills/loop-dog-food/SKILL.md

loop-dog-food

Skill0 starsChanged 5 months agoMITView on GitHub

Turn a spec, plan, orchestration pattern, or other work object into a bounded loop with visible state, a turn budget, review gates, and proof commands.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when the operator wants the repo, spec, plan, or orchestration pattern to improve itself through the loop.

    line 8
  2. Distinctive:

    Treat the work object as the current goal surface.

    line 23
  3. Very specific:

    Keep permission boundaries visible in loop-state.md.

    line 28
  4. Distinctive:

    Stop if the loop needs destructive action, hidden credentials, broader permissions, or proof that cannot be run.

    line 29
  5. Very specific:

    Update loop-state.md with the chosen task, proof command, stop condition, and next action.

    line 42

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Loop Dog Food

Use this skill when the operator wants the repo, spec, plan, or orchestration pattern to improve itself through the loop.

Inputs

  • loop-state.md
  • loop-contract.md
  • loop-decisions.md
  • loop-prompts.md
  • A named work object, such as a spec, plan, branch, issue, repo area, or orchestration pattern
  • A turn budget: single, multi, or infinite

Process

  1. Treat the work object as the current goal surface.
  2. Zoom out enough to identify the highest-leverage next loop task.
  3. Choose exactly one bounded task unless the turn budget explicitly permits additional iterations.
  4. Name the proof command before implementation starts.
  5. Keep permission boundaries visible in loop-state.md.
  6. Stop if the loop needs destructive action, hidden credentials, broader permissions, or proof that cannot be run.

Turn Budgets

  • single: complete one bounded iteration, record it, then stop.
  • multi: continue while proof stays green and the state board names a next action.
  • infinite: continue while compute and permission boundaries allow, but never bypass review, proof, or stop conditions.

Output

Update loop-state.md with the chosen task, proof command, stop condition, and next action. Record durable decisions in loop-decisions.md.

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.