agentleFS
Sign inStart free
MasihMoafiMasihMoafi/skills/project-docs/SKILL.md

project-docs

Skill8 starsChanged 9 days agoMITView on GitHub

Use when entering an unfamiliar project, when creating or repairing a project's agent docs (AGENTS.md, feature map, lat.md folder map, VISION.md, ES.md), or when the workspace check reports doc problems.

22 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use to-issues, to-prd and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • GoogleSkill

    to-issues

    Break a plan, spec, or PRD into independently-grabbable issues on the project issue tracker using tracer-bullet vertical slices.

    Also in 3 projects that use this

  • GoogleSkill

    to-prd

    Turn the current conversation into a PRD and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.

    Also in 3 projects that use this

  • Google CloudSkill

    triage

    Gather context on a change or bug and plan the implementation.

    Also in 3 projects that use this

  • BlockSkill

    frontend-design

    Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 18 edited versions we read keep these.

  • 2

    Distinctive

    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
  1. Distinctive:

    When a doc contradicts the code or another doc, check the code, fix or delete the wrong text, and say what changed.

    line 42
  2. Distinctive:

    Before you delete old text, move each lesson or prohibition that is still valid to a feature's gotchas or to AGENTS.md.

    line 43
  3. Very specific:

    Add a harness part (init.sh, a rubric, a lint) only for a workflow that repeats, and only the smallest that fixes the bottleneck.

    line 44

“Common” is measured: kept by at least 6 in 10 of the 18 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Project docs

Each fact about a project has one home. Read in this order: AGENTS.md (loaded already), ES.md, then the feature file for the task. Use the folder map before you add a file and VISION.md when scope is in question. A README supplements these files; it never replaces them.

One file per job

FileHoldsLimit
AGENTS.md (+ CLAUDE.md = @AGENTS.md)## Start here (pointers), ## Common mistakes (this project's lessons), ## Rules60 lines
ES.md, title "(updated YYYY-MM-DD)"## Now, ## Next, ## Broken, ## Open decisions40 lines
VISION.mdidentity in one line, ## Thesis, ## Non-goals25 lines
lat.md/workspace.mdevery top-level item and why it exists—
.agents/skills/verify-<app>/features/one file per user-facing feature: how to reach it, drive it, its code, its gotchas; made by create-verification-skill, kept true by maintain-verification-skill—
reports/, commit messagesevidence: hashes, run IDs, verification prose—

A workspace check can enforce the limits and headings (here: findability/check_pointers.py). These files are local notes: keep them out of public git and link them into new worktrees.

What goes in AGENTS.md

It loads into every session, so each line must apply to every task there (HumanLayer, Writing a Good CLAUDE.md; Anthropic, Steering Claude Code).

  • Yes: what the project is, how to build, test and verify it, where things live (as pointers), and rules that hold for every task.
  • No: procedures (a skill), task-specific notes (the feature file), state (ES.md), copies of global rules (the user-level instructions), code style (a linter).
  • A "never do X" that agents keep breaking needs a check or hook; the line in AGENTS.md is not enough.
  • A common mistake is one dated line: what went wrong, the rule, and its guard (or "No guard").

What goes in ES.md

  • ## Now: what is true today, in up to five lines. Each line states a fact; its proof goes to reports/.
  • ## Next: the user's open requests, newest date first, one line each, starting Feature:, Fix:, Remove: or Keep:. Delete a line only when the user has seen it working. When the log outgrows the file, it moves to a local REQUESTS.md beside ES.md (same line format), and ## Next keeps the user's ranked items and a link.
  • ## Broken and ## Open decisions: one line each; a decision names who decides.
  • Rewrite the file in place; never append a progress log.

Keep them true

  • Change a fact in its home only; elsewhere, point to it.
  • When a doc contradicts the code or another doc, check the code, fix or delete the wrong text, and say what changed.
  • Before you delete old text, move each lesson or prohibition that is still valid to a feature's gotchas or to AGENTS.md.
  • Add a harness part (init.sh, a rubric, a lint) only for a workflow that repeats, and only the smallest that fixes the bottleneck. If you are not sure it is still needed, remove it and keep the simpler setup if quality holds.

Create the layout when absent

Inspect the project two levels deep, draft the missing files in the shape above, and show the user VISION.md before you save it. Exemplar: this library's VISION.md.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free