meta-skills
AGENTS.md14 starsChanged 3 months agoMITView on GitHub
Before project work, read the PROGRESS snapshot and STRUCTURE's "Read before changing" section. Read the other records when the task needs them. Use the horsemen skill to update them. - This repository holds portable skills and plugin metadata. Hosts supply tools, permissions, identity, memory, and continuity. No runtime is required. - Preserve each skill's contract. The default skill uses the smallest relevant capability set. Standalone outcome skills remain usable on their own. - Use the customer's brand. Never invent capabilities,…
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 31,700 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 repository22 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- brief-creativeSkill
- brief-product-uiSkill
- conquistadorSkill
- create-brandSkill
- create-paid-campaignSkill
- create-shortformSkill
- debate-agentsSkill
- improve-conversionSkill
- knowledge-reviewSkill
- map-user-flowSkill
- measure-growthSkill
- model-growth-funnelSkill
- optimize-searchSkill
- plan-campaignSkill
- polish-vietnameseSkill
- research-channelSkill
- research-positioningSkill
- write-copySkill
- write-longformSkill
- write-outreachSkill
- write-socialSkill
The file
We split this AGENTS.md into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is 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.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 3
Before project work, read the PROGRESS snapshot and STRUCTURE's "Read before changing" section.
- Distinctive: line 18
Use the customer's brand. Never invent capabilities, proof, metrics, or consent.
- Distinctive: line 19
Keep consequential actions and persistence behind the approval gates in PROGRESS.md.
- Distinctive: line 23
Keep private data, identifiers, account emails, and machine paths out of this public repository.
≈ “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
Project context
Before project work, read the PROGRESS snapshot and STRUCTURE's
"Read before changing" section. Read the other records when the task needs
them. Use the horsemen skill to update them.
- VISION.md: draft direction, assumptions, and owner questions
- BRAIN.md: knowledge locations, save rules, and access limits
- PROGRESS.md: current state, gates, and next steps
- STRUCTURE.md: package map and required reading
Package rules
- This repository holds portable skills and plugin metadata. Hosts supply tools, permissions, identity, memory, and continuity. No runtime is required.
- Preserve each skill's contract. The default skill uses the smallest relevant capability set. Standalone outcome skills remain usable on their own.
- Use the customer's brand. Never invent capabilities, proof, metrics, or consent.
- Keep consequential actions and persistence behind the approval gates in PROGRESS.md. If the host cannot enforce approval, return a draft.
- A skill's frontmatter version is authoritative. Keep VERSIONS.md in sync without forcing skill versions to equal the package version.
- Keep private data, identifiers, account emails, and machine paths out of this public repository. Follow BRAIN.md.
- Existing installation instructions are not proof of a tested install. Record the evidence and its original date when you report delivery.
Inherited higher-priority instructions continue to apply.
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.