agentleFS
Sign inStart free
fasterquieterfasterquieter/AgentOS-Public/AGENTS.md

AgentOS-Public

AGENTS.md0 starsChanged 55 days agoMITView on GitHub

This repository is both the AgentOS framework project and a complete seed for a different project. Resolve that identity before loading project context. Read .agentos/seed-state.md. - If the user's substantive objective is to build or maintain AgentOS itself, this is the master framework checkout. Follow Master mode below. - If the seed is unadapted and the user supplies a substantive specification for any other product, that ordinary product prompt triggers the one-time .agentos-kit/ADAPTATION.md protocol. Do not ask the user to…

  • 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 31,774 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 repository2 other files it gives its agents.

The file

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

9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    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
  1. Distinctive:

    If the user's substantive objective is to build or maintain AgentOS itself, this is the master framework checkout.

    line 10
  2. Very specific:

    If the seed is unadapted and the user supplies a substantive specification for any other product, that ordinary product prompt triggers the …

    line 12
  3. Very specific:

    Before resolving the gate, do not read the current .agentos/ tree, framework history, examples, or .agentos-kit/ beyond the adaptation file …

    line 21
  4. Distinctive:

    Keep root AGENTS.md a router.

    line 32
  5. Distinctive:

    Never claim behavior from compilation alone.

    line 38
  6. Very specific:

    Treat .agentos-kit/ as reusable framework source, not AgentOS project state.

    line 40
  7. Distinctive:

    Treat its result as recommendations.

    line 50
  8. Distinctive:

    Update an active plan before stopping unfinished substantial work.

    line 65
  9. Very specific:

    Put a general framework lesson in .agentos/upstream/ only with project evidence and explicit craftsmanship/simplification review.

    line 68

≈ “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.

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

AgentOS master and project-seed router

This repository is both the AgentOS framework project and a complete seed for a different project. Resolve that identity before loading project context.

Identity gate — do this first

Read .agentos/seed-state.md.

  • If the user's substantive objective is to build or maintain AgentOS itself, this is the master framework checkout. Follow Master mode below.
  • If the seed is unadapted and the user supplies a substantive specification for any other product, that ordinary product prompt triggers the one-time .agentos-kit/ADAPTATION.md protocol. Do not ask the user to initialize AgentOS, choose modules, run commands, or restate the prompt. Adapt the clone, then continue into product work in the same task.
  • Housekeeping with no new product specification does not trigger adaptation.

The task subject is decisive. Directory name, inherited Git history, and a copied AgentOS remote do not make a non-AgentOS product into AgentOS work. Before resolving the gate, do not read the current .agentos/ tree, framework history, examples, or .agentos-kit/ beyond the adaptation file it names.

Master mode

AgentOS is Markdown-first institutional memory plus optional deterministic helpers. It is not an agent runtime or a documentation dump.

Always

  • Inspect Git state before editing and preserve unrelated work.
  • Keep root AGENTS.md a router; durable detail belongs in indexed memory.
  • Honor effective catalog change_control; an agent cannot weaken a governing policy to authorize its own edit.
  • Preserve .agentos/invariants.md.
  • Separate observed evidence, accepted decisions, working hypotheses, and history.
  • Never claim behavior from compilation alone.
  • Replace obsolete current truth instead of appending another version.
  • Treat .agentos-kit/ as reusable framework source, not AgentOS project state.

Find context

Inspect .agentos/index.json. The optional local finder is:

PYTHONPATH=.agentos-kit/src python3 -m agentos context "<task>" --path <target-path>

Treat its result as recommendations. Read relevant documents in order, inspect reported omissions selectively, and never read the full memory or capability kit by default.

Development

  • Runtime: Python 3.10+, standard library only.
  • Tests: PYTHONPATH=.agentos-kit/src python3 -m unittest discover -s .agentos-kit/tests -v.
  • Validate: PYTHONPATH=.agentos-kit/src python3 -m agentos validate.
  • Public behavior changes require tests and README/help reconciliation.
  • Seed changes require complex, tiny, later-activation, fresh-handoff, and context evaluations.

Closeout

  • Update an active plan before stopping unfinished substantial work.
  • Reconcile durable discoveries into current knowledge or a decision.
  • Archive/delete absorbed transient handoffs; Git holds implementation history.
  • Put a general framework lesson in .agentos/upstream/ only with project evidence and explicit craftsmanship/simplification review.

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