project-profile
Skill5 starsChanged 14 days agoMITView on GitHub
The project profile (stage, audience, team, sensitive data) and which practices it calls for, so the engineering fits this project instead of a large enterprise. Use when deciding how much rigor a change needs, when a skill item carries a profile tag such as (production), when proposing a profile in a project scan, or when a design feels heavier than the project.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- brilliantrough/agent-skillstheir own version, first seen Nov 8, 2026
- juanitourquiza/shipframetheir own version, first seen Oct 23, 2026
- awahids/noskilltheir own version, first seen Oct 15, 2026
- lucastosetto/dev-skillstheir own version, first seen Oct 15, 2026
- analyticsengineering/ae-toolkittheir own version, first seen Oct 12, 2026
- jlrrequina/pi-project-profiletheir own version, first seen Sep 29, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- brilliantrough/agent-skills★ 11same name
- juanitourquiza/shipframe★ 4same name
- analyticsengineering/ae-toolkit★ 3same name
- awahids/noskill★ 1same name
- jlrrequina/pi-project-profile★ 0same name
- lucastosetto/dev-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository11 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- commits-and-prsSkill
- context-mdSkill
- debuggingSkill
- diagramsSkill
- implementationSkill
- project-scanSkill
- session-retroSkill
- task-plansSkill
- tddSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
3
Distinctive
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 10
Not every project needs every practice.
- Very specific: line 21
If the profile is missing, assume early, internal, small, none, and say so.
- Distinctive: line 38
Test the behavior you change, and never weaken a test.
- Distinctive: line 39
Stop before anything that loses data or can't be undone, and never apply infrastructure changes or deploy yourself.
- Very specific: line 44
Stage
prototype: test-first for logic you will keep.
“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- From agentic-defaults. Kit updates replace this file except a "## This project" section at the end, which belongs to the project. Edit nothing else. Rules for all work go in AGENTS.md "This project". -->Project profile
Not every project needs every practice. The profile in the "This project" section of AGENTS.md says what kind of project this is, and that decides which costly practices apply. Cheap correctness and safety always apply.
The four dimensions
| Dimension | Values |
| Stage | prototype (exploring, may be thrown away), early (first real users), production (people depend on it), mature (long-lived, many dependents) |
| Audience | personal, internal (your own organization), customer (specific clients, often under contract), public (anyone on the internet) |
| Team | solo, small (one team), several (teams depending on each other's code) |
| Handles | none, or any of money, personal data, regulated data, safety |
If the profile is missing, assume early, internal, small, none, and say so.
Tags
Items in the kit's skills marked with a tag apply only when the profile matches. Items without a tag always apply. An item with several tags applies when any one of them matches, and a tag on a heading covers everything under it.
| Tag | Applies when |
(production) | Stage is production or mature |
(external) | Audience is customer or public |
(public) | Audience is public |
(teams) | Team is several |
(sensitive) | Handles is anything but none |
Always, whatever the profile
- Validate input, check access to the specific record, use parameterized queries, keep secrets out of code.
- Test the behavior you change, and never weaken a test.
- Stop before anything that loses data or can't be undone, and never apply infrastructure changes or deploy yourself.
- End every task with the context check.
What each value adds or relaxes
- Stage
prototype: test-first for logic you will keep; glue code and layout may go untested. Breaking changes are fine. A rootCONTEXT.mdis enough. Say which code is throwaway. - Stage
early: the kit's defaults, without the tagged items. (production): migrations that keep live data and running code working, no breaking change without a migration path, and switches for unfinished features.(external): versioned, documented API contracts; contract tests for what clients use; a security review for every change to authentication, permissions or input handling.(public): also abuse limits (rate limits, input size caps) and accessibility checks.(teams): decision records for choices other teams depend on; contracts first, with contract tests, between teams.(sensitive): a security review for changes that touch that data, an audit trail of who changed it, and a backup before migrations that change it.
Don't build for a bigger project
Skip a practice whose cost outweighs what it prevents here, and say so in one line. Signs of building for a bigger project than this one:
- layers, interfaces or base classes with a single implementation,
- generic repositories, event buses, plugin systems, configuration frameworks,
- separate services, queues or infrastructure that no requirement asks for,
- practices tagged for a profile this project doesn't have.
Proposing a profile
In a project scan, propose a profile from evidence, mark it "(proposed)", and ask the user to confirm it:
- Stage: deployment files, deploy steps in CI, release tags, a live URL in the README.
- Audience: how users sign in, public endpoints, client names in the README or tickets.
- Team: the survey's contributor count, a
CODEOWNERSfile. - Handles: payment, billing or rate fields, personal data in the models, compliance notes.
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.