skills-marketplace / templates
AGENTS.md1 starsChanged 36 days agoMITView on GitHub
Every dev-flow skill uses only these. - Start from the ticket and its Files. Search before opening; open files one at a time. - Never bulk-read the repo, lockfiles, node_modules, generated code or fixtures (first 40 lines at most). - Never read or print .env*. Never edit AGENTS.md or .cursor/rules/* during a ticket; propose changes in the report. - Shortest complete answer. No restating the task, no pasted diffs. <!-- dev-flow:end -->
- Reads credentials
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 32,671 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 repository14 other files it gives its agents.
- grill-meSkill
- addressSkill
- conventional-commitsSkill
- design-packageSkill
- doneSkill
- implementSkill
- preview-prSkill
- researchSkill
- setup-projectSkill
- stop-previewSkill
- takeSkill
- ticketsSkill
- ui-from-referenceSkill
- visual-notesSkill
The file
We split this AGENTS.md into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Every dev-flow skill uses only these.
- Distinctive: line 47
Start from the ticket and its Files.
- Very specific: line 48
Never bulk-read the repo, lockfiles, node_modules, generated code or fixtures (first 40 lines at most).
- Very specific: line 49
Never read or print .env*.
≈ “Common” is an estimate: at least 1 in 5 of the 41,581 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_name>>
<<one_sentence>>
<!-- dev-flow:start (written by /setup-project; edit freely, keep this block under 60 lines) -->Stack
<<stack_line>>
Where things are
| Path | What |
| <<paths_rows>> |
Commands
Every dev-flow skill uses only these.
| Command | Does |
just test-q | tests, quiet; prints TRUNCATED if output was cut |
just typecheck | static checks |
just lint | lint and format check |
just check | typecheck + lint + tests; must pass before a PR |
just run / just stop | start the app in the background / stop it (just logs to watch) |
just task <n> <todo|in-progress|in-review|done> | move a ticket (label + board) |
| <<extra_command_rows>> |
Work
- Tickets: GitHub Issues in
<<repo>>, board "<<board_title>>", one milestone per phase. Seedocs/agents/issue-tracker.md. - Spec:
<<spec_path>>. Decisions and lessons:docs/DECISIONS.md. Design:docs/design/README.md. - Git: <<git_tool>>, base branch
<<base_branch>>. The human merges.
Models
| Work | Model |
| /tickets, /take | <<model_plan>> |
| /implement, UI translation | <<model_build>> |
| /address, /done | <<model_light>> |
| Stuck after two attempts | <<model_stuck>> |
Always
- Start from the ticket and its Files. Search before opening; open files one at a time.
- Never bulk-read the repo, lockfiles,
node_modules, generated code or fixtures (first 40 lines at most). - Never read or print
.env*. Never editAGENTS.mdor.cursor/rules/*during a ticket; propose changes in the report. - Shortest complete answer. No restating the task, no pasted diffs.
<!-- dev-flow:end -->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.

