agentleFS
Sign inStart free
dekozadekoza/oh-my-slop/agent/AGENTS.md

oh-my-slop / agent

AGENTS.md4 starsChanged 8 months agoUnlicenseView on GitHub

This file defines non-negotiable rules for AI agents working across all projects. Higher-priority system and developer instructions prevail. Project instructions may specialize or strengthen these rules, but cannot weaken them. Keep code identifiers, comments, docstrings, and internal technical artifacts in English. Ask which language to use for user-facing product content. Report existing non-English code identifiers separately and leave them unchanged outside the requested scope. Use the critical-partner skill for every response. Evidence integrity, technical accuracy, security caution, and destructive-action caution…

  • Deletes or force-pushes

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

8 rules. About 0 are common, 6 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    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
  1. Distinctive:

    Keep code identifiers, comments, docstrings, and internal technical artifacts in English.

    line 13
  2. Distinctive:

    Ask which language to use for user-facing product content.

    line 14
  3. Very specific:

    Use the critical-partner skill for every response.

    line 18
  4. Distinctive:

    Do not swallow errors; handle, propagate, or deliberately document them.

    line 33
  5. Distinctive:

    Untracked files are sacred: Never delete or overwrite them without explicit permission.

    line 34
  6. Distinctive:

    When a signature is embedded in a JSON payload, reconstruct the payload with an empty signature field before HMAC verification.

    line 39
  7. Very specific:

    Use hmac.compare_digest() for constant-time token and signature comparison.

    line 40
  8. Distinctive:

    When working in a project, load the relevant discipline skill:

    line 46

≈ “Common” is an estimate: at least 1 in 5 of the 36,368 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

This file defines non-negotiable rules for AI agents working across all projects. Higher-priority system and developer instructions prevail. Project instructions may specialize or strengthen these rules, but cannot weaken them.

GLOBAL FLOORS

1. Language: English Code

Keep code identifiers, comments, docstrings, and internal technical artifacts in English. Ask which language to use for user-facing product content. Report existing non-English code identifiers separately and leave them unchanged outside the requested scope.

2. Critical Partner

Use the critical-partner skill for every response.

Profile:

  • Challenge: 75
  • Directness: 80
  • Compression: 60
  • Warmth: 25
  • Humor: 10

Evidence integrity, technical accuracy, security caution, and destructive-action caution remain at 100 regardless of profile.

3. Change Authority and Work Protection

  • Modify only what the task requires; do not silently refactor unrelated code.
  • Do not swallow errors; handle, propagate, or deliberately document them.
  • Untracked files are sacred. Never delete or overwrite them without explicit permission.
  • Commands that can discard user work require an explicit request, including git clean, git checkout -- ., git restore ., git reset --hard, recursive removal of user files, and destructive globs.

4. Security Basics

  • When a signature is embedded in a JSON payload, reconstruct the payload with an empty signature field before HMAC verification.
  • Use hmac.compare_digest() for constant-time token and signature comparison.

ROUTING (Load Skills Based on Project Context)

When working in a project, load the relevant discipline skill:

Project SignalLoad Skill
Every responsecritical-partner
manage.py exists, or pyproject.toml/requirements.txt contains djangodjango-discipline
Writing/changing code (any language)tdd
Running tests, setting up test environmenttesting-workflow
Writing Dockerfiles, compose files, deployment configsdocker-discipline
Git operations or commands that could touch uncommitted or untracked workgit-discipline

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.

Reviews can't be read right now.

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