skills
AGENTS.md3 starsChanged 3 months agoMITView on GitHub
This repository teaches agents how to think about systems. It recruits knowledge through generative terms and reasoning moves instead of trying to contain every relevant fact. One skill closes one behavioral gap. Prefer a leading term that activates the right prior knowledge over an inventory of everything the agent might inspect. Let the decision govern the depth. Flexible reasoning is the default. Stable contracts are welcome where shared reliance needs them. The method applies across domains. Software is one example,…
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,542 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository21 other files it gives its agents.
- automating-systemsSkill
- changing-systemsSkill
- choosing-interventionsSkill
- creating-skillsSkill
- designing-interfacesSkill
- domain-modelingSkill
- evaluating-systemsSkill
- framing-decisionsSkill
- governing-systemsSkill
- implementing-systemsSkill
- making-videosSkill
- operating-a-systemSkill
- prototypeSkill
- representing-systemsSkill
- reviewSkill
- setup-system-thinkingSkill
- thinking-in-systemsSkill
- to-specSkill
- to-ticketsSkill
- transition-handoffSkill
- wayfinderSkill
The file
We split this AGENTS.md into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 21
Put stable semantics in a standard.
- Very specific: line 33
Follow the information hierarchy and single-source rule from writing-for-agents.
- Very specific: line 35
Keep shared sources in SOURCES.md.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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.
Files
Systems thinking skills
This repository teaches agents how to think about systems. It recruits knowledge through generative terms and reasoning moves instead of trying to contain every relevant fact.
What matters here
Teach judgment
One skill closes one behavioral gap. Prefer a leading term that activates the right prior knowledge over an inventory of everything the agent might inspect.
Stay proportional
Let the decision govern the depth. Flexible reasoning is the default. Stable contracts are welcome where shared reliance needs them.
Transfer across domains
The method applies across domains. Software is one example, not the default model.
Keep standards adaptable
Put stable semantics in a standard. Put project-specific choices in a disclosed reference that can override the default without redefining the semantics.
Shared language
The Thinking in Systems skill and its context govern every skill in this repository. Read both before creating, editing, or reviewing a skill.
Reuse the glossary's terms instead of restating their definitions. Add a cross-cutting term to the glossary before introducing a synonym. A focused skill may narrow a term for its branch, but it must preserve the shared meaning.
Authoring
Apply writing-for-agents and writing-guidelines to every agent-consumed document. For skill work, also apply creating-skills and read writing-for-agents/SKILL-MECHANICS.md.
Follow the information hierarchy and single-source rule from writing-for-agents. Show the target behavior with a concrete case when prose leaves room for imitation without understanding.
Keep shared sources in SOURCES.md. Focused skills point to it instead of owning source catalogs.
SKILL.md is the source of truth for agents/openai.yaml. After editing a skill, run node scripts/generate-openai-yaml.mjs, then verify with node scripts/generate-openai-yaml.mjs --check.
Edit the personal opening of README.md and the “Note from Yannic” only with Yannic in the loop.
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.