owner-operator
AGENTS.md8 starsChanged 44 days agoMITView on GitHub
This file owns policy. Skills own workflows. Behavior knowledge lives in docs/: run npm run docs:list, then read the relevant pages before working on a surface. Read a subtree's AGENTS.md before working in it. - Be concise. High signal, low noise, always. Digestible information promotes decision-making. - Cite causal claims granularly. Hand-waved conclusions are dangerous, and the reader will ask for sources anyway. - Don't reinvent. Proven, maintained open source beats local code. Start at docs/inspiration.md, then search (gh-search skill);…
In this repository75 other files it gives its agents.
- docs/AGENTS.mdAGENTS.md
- workspace/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- ask-mattSkill
- codebase-designSkill
- code-reviewSkill
- diagnosing-bugsSkill
- domain-modelingSkill
- gh-searchSkill
- grillingSkill
- grill-meSkill
- grill-with-docsSkill
- handoffSkill
- implementSkill
- improve-codebase-architectureSkill
- prototypeSkill
- pstack-architectSkill
- pstack-arenaSkill
- pstack-automate-meSkill
- pstack-blast-radiusSkill
- pstack-broSkill
- pstack-create-verification-skillSkill
- pstack-figure-it-outSkill
- pstack-howSkill
- pstack-interrogateSkill
- pstack-maintain-verification-skillSkill
- pstack-make-bot-uiSkill
- pstack-no-commentsSkill
- pstack-poteto-modeSkill
- pstack-principle-attack-the-premiseSkill
- pstack-principle-boundary-disciplineSkill
- pstack-principle-build-the-leverSkill
- pstack-principle-encode-lessons-in-structureSkill
- pstack-principle-exhaust-the-design-spaceSkill
- pstack-principle-experience-firstSkill
- pstack-principle-fix-root-causesSkill
- pstack-principle-foundational-thinkingSkill
- pstack-principle-guard-the-context-windowSkill
- pstack-principle-laziness-protocolSkill
- pstack-principle-make-operations-idempotentSkill
- pstack-principle-migrate-callers-then-delete-legacy-apisSkill
- pstack-principle-minimize-reader-loadSkill
- pstack-principle-model-the-domainSkill
- pstack-principle-never-block-on-the-humanSkill
- pstack-principle-outcome-oriented-executionSkill
- pstack-principle-prove-it-worksSkill
- pstack-principle-redesign-from-first-principlesSkill
- pstack-principle-separate-before-serializing-shared-stateSkill
- pstack-principle-sequence-verifiable-unitsSkill
- pstack-principle-subtract-before-you-addSkill
- pstack-principle-test-behavior-not-implementationSkill
- pstack-principle-type-system-disciplineSkill
- pstack-recallSkill
- pstack-reflectSkill
- pstack-setup-pstackSkill
- pstack-show-me-your-workSkill
- pstackSkill
- pstack-swarmSkill
- pstack-tddSkill
- pstack-teachSkill
- The first 60 shown.
The file
Files
Owner Operator — Development
This file owns policy. Skills own workflows. Behavior knowledge lives in
docs/: run npm run docs:list, then read the relevant pages before
working on a surface. Read a subtree's AGENTS.md before working in it.
- Be concise. High signal, low noise, always. Digestible information promotes decision-making.
- Cite causal claims granularly. Hand-waved conclusions are dangerous, and the reader will ask for sources anyway.
- Don't reinvent. Proven, maintained open source beats local code. Start at
docs/inspiration.md, then search (gh-search skill);
cite what you keep to pinned lines (
…/blob/<SHA>/path#L10-L40). - Documentation lives in one place. Pointers survive change; restated behavior silently goes stale.
- Complexity must earn its keep. It's almost always better to build up than tear down. Challenge complexity.
- Durable text must make sense without prior context. Docs, comments, and prompts outlive the conversation that produced them; reread as a stranger before saving.
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.

