loop-builder / templates
AGENTS.md2 starsChanged 2 months agoMITView on GitHub
The operational how. <60 lines. Ratchet: every line traceable to a real failure or a hard constraint. NOT a changelog or progress file (that's IMPLEMENTATION_PLAN.md).
In this repositoryOne other file it gives its agents.
- loop-builderSkill
The file
Files
AGENTS.md — {{TASK_TITLE}}
The operational how. <60 lines. Ratchet: every line traceable to a real failure or a hard constraint. NOT a changelog or progress file (that's IMPLEMENTATION_PLAN.md).
Loop engine
{{ENGINE}} <!-- loop.sh | /ralph-loop | subagents (Task tool) | OS scheduler -->
Commands (backpressure / eval-gate)
- Build: {{BUILD_CMD}}
- Test: {{TEST_CMD}}
- Lint: {{LINT_CMD}}
- Run: {{RUN_CMD}}
Where the code / output lives
{{SRC_LOCATION}}
Operating rules (ratchet — append one failure at a time)
- Surgical changes. Nothing out of scope.
- Irreversible/external actions (deploy, push to prod, email/DM, delete) → confirm with the human, never auto.
- One task per iteration. Validate BEFORE committing.
Learnings (discovered during the loop)
<!-- The agent appends gotchas/conventions discovered here, so later iterations benefit. -->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
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.