agentleFS
akovfoj-langakovfoj-lang/plainframe/AGENTS.md

plainframe

AGENTS.md0 starsChanged 23 days agoMITView on GitHub

Identical rules for every agent, whichever tool you are (Claude, Codex/GPT, anything future). There is no separate rulebook. One term recurs everywhere below: the owner is the human running this repo — as opposed to you, the agent operating it. Every "the owner" in these docs means that person, never you. 1. Run os/scripts/session-start.sh first — before reading CLAUDE.md, so the constitution you're about to read can't be stale (law 8). It pulls (both ways safe: it never commits, pushes,…

  • 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 repository11 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    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
  1. Very specific:

    Run os/scripts/session-start.sh first — before reading CLAUDE.md, so the constitution you're about to read can't be stale (law 8).

    line 10

≈ “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.

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

AGENTS.md — start here

Identical rules for every agent, whichever tool you are (Claude, Codex/GPT, anything future). There is no separate rulebook.

One term recurs everywhere below: the owner is the human running this repo — as opposed to you, the agent operating it. Every "the owner" in these docs means that person, never you.

  1. Run os/scripts/session-start.sh first — before reading CLAUDE.md, so the constitution you're about to read can't be stale (law 8). It pulls (both ways safe: it never commits, pushes, or touches uncommitted work), lists the newest receipts, and names stuck work. Claude Code runs it for you at startup. If a later pull in this session changes CLAUDE.md (e.g. during a sync), re-read it in full before continuing — "read once" means once per fresh copy, not once no matter what the tree does underneath you.
  2. Read CLAUDE.md next, then MAP.md, and follow only the pages MAP routes you to.
  3. Sync your own work yourself, after each finished unit and before you stop: os/scripts/sync.sh "<what you did>" --receipt "<topic>: <one line>" -- <your paths>.

Tool-specific command adapters live in .claude/commands/ (Claude Code, native) and .agents/skills/ (Codex, native — other agents can still run Plainframe by reading the playbooks in os/playbooks/ directly, just without a generated command surface yet; see README's "One repo, many agents"). Both generated sets come from os/commands.md — the playbooks are the single source of truth. If your adapters and the manifest disagree, regenerate: os/scripts/gen-commands.sh.

Claude Code also runs two hooks of its own (.claude/settings.json): os/scripts/claude-hooks/tool-guard.sh asks before an edit to a protected path, and outside-content.sh flags an approval token or injection phrasing in what a fetch brings back (law 10). Deny rules in the same file refuse the Never tier (force-push, skipping the git hooks, reading .env). They bind Claude Code only; the git hooks bind every agent.

The kernel holds nothing tool-specific. Same repo, same laws, interchangeable operators.

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.