SreeStack
AGENTS.md1 starsChanged 16 days agoMITView on GitHub
This repo is the source of truth for user skills. Edit them here, not in live skill folders. Built-in and plugin-managed skills stay with their hosts. - Design each skill as one coherent tool for its current task. On every revision, especially a large one, reconsider its structure as if writing it from scratch. Preserve useful behavior and important details, not inherited file layouts or wording. Merge overlap and resolve conflicting guidance. - Organize references by the task, concept, or…
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 31,843 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 repository57 other files it gives its agents.
- agent-loginSkill
- animateSkill
- app-intentsSkill
- app-store-connectSkill
- architectSkill
- arenaSkill
- blast-radiusSkill
- codebase-designSkill
- code-qualitySkill
- correctSkill
- diagnosing-bugsSkill
- domain-modelingSkill
- eli5Skill
- engineering-principlesSkill
- gauntlet-loopSkill
- gh-stackSkill
- goal-swarmSkill
- grillingSkill
- grill-with-docsSkill
- handoffSkill
- howSkill
- improve-codebase-architectureSkill
- ios-hapticsSkill
- koubouSkill
- mergeSkill
- orchestrationSkill
- prototypeSkill
- prSkill
- react-best-practicesSkill
- react-doctorSkill
- reframeSkill
- repo-cleanupSkill
- researchSkill
- retroSkill
- review-fix-loopSkill
- reviewSkill
- review-sweepSkill
- seoSkill
- setup-sreestackSkill
- show-dont-tellSkill
- showroomSkill
- snipsnap-do-choresSkill
- snip-snapSkill
- swift-testing-modernizationSkill
- swiftuiSkill
- tddSkill
- to-specSkill
- to-ticketsSkill
- triageSkill
- ui-designSkill
- uikitSkill
- verifySkill
- wayfinderSkill
- web-component-inspirationSkill
- whySkill
- writing-for-agentsSkill
- xcode-security-auditSkill
The file
We split this AGENTS.md into the rules it gives an agent
6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 13
Keep origins, import history, and source mappings in SOURCES.md.
- Very specific: line 17
Keep skills/ flat. Put each skill’s supporting files inside its folder.
- Distinctive: line 18
Use current local skill names in source records.
- Very specific: line 20
Before modifying a skill, read its entry in SOURCES.md for its import baseline and local differences, including invocation choices.
- Very specific: line 25
Update SOURCES.md when an origin, import baseline, or local difference changes.
- Distinctive: line 28
Keep skill instructions, setup, and supporting references inside each skill's folder.
≈ “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
SreeStack
This repo is the source of truth for user skills. Edit them here, not in live skill folders. Built-in and plugin-managed skills stay with their hosts.
- Design each skill as one coherent tool for its current task. On every revision, especially a large one, reconsider its structure as if writing it from scratch. Preserve useful behavior and important details, not inherited file layouts or wording. Merge overlap and resolve conflicting guidance.
- Organize references by the task, concept, or interaction they teach. Link them
from
SKILL.mdwhere an agent needs them. Source authors and upstream skill boundaries must not shape the skill's instructions or reference groups. - Keep origins, import history, and source mappings in
SOURCES.md; put credits inREADME.mdand required license notices in the repo's notice files. Skill instructions should stand on their own without origin notes. Keep technical citations beside guidance when they support an API or behavior claim. - Keep
skills/flat. Put each skill’s supporting files inside its folder. - Use current local skill names in source records. Preserve upstream names and paths that identify imported sources.
- Before modifying a skill, read its entry in
SOURCES.mdfor its import baseline and local differences, including invocation choices. - Upstream updates must preserve local adaptations and invocation choices unless the user asks to change them. The recorded revision is the import baseline, not the latest version checked.
- Update
SOURCES.mdwhen an origin, import baseline, or local difference changes. Describe the current difference, not how the skill was built. Keep local-only skills to a short origin note. Git holds edit, rename, and review history. - Keep skill instructions, setup, and supporting references inside each skill's folder. Use README.md for the repository overview, skill discovery, and links.
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.