agentleFS
Sign inStart free
anotbanotb/x-signal/AGENTS.md

x-signal

AGENTS.md1 starsChanged 3 months agoMITView on GitHub

  • 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 31,700 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 repository4 other files it gives its agents.

The file

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

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

  • ≈2

    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 the default services and ports bound to loopback.

    line 6
  2. Very specific:

    Start with docker compose up -d --build --wait, then check /healthz, /readyz, and docker compose ps.

    line 7
  3. Common:

    Never ask a user to paste cookies, session tokens, VNC passwords, tunnel URLs, or other credentials into chat.

    line 9
  4. Distinctive:

    Prefer a dedicated Chrome profile with the session companion.

    line 10
  5. Distinctive:

    Do not bypass X authentication, challenges, access controls, or rate limits.

    line 11
  6. Distinctive:

    Keep the architecture tool-only unless a real user workflow requires UI.

    line 17
  7. Distinctive:

    Do not add posting, generic browser automation, CAPTCHA handling, proxy rotation, or enforcement-evasion features.

    line 19
  8. Common:

    Use synthetic parser fixtures.

    line 20
  9. Very specific:

    Run npm run validate, npm audit --audit-level=moderate, docker compose config --quiet, and the plugin/skill validators before release.

    line 21
  10. Distinctive:

    Treat live tests as real account traffic.

    line 22

≈ “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

Working with X Signal

Installing for a user

  • Detect the operating system and confirm Docker Compose is available before changing anything.
  • Keep the default services and ports bound to loopback.
  • Start with docker compose up -d --build --wait, then check /healthz, /readyz, and docker compose ps.
  • Connect local Codex clients to http://127.0.0.1:7345/mcp.
  • Never ask a user to paste cookies, session tokens, VNC passwords, tunnel URLs, or other credentials into chat.
  • Prefer a dedicated Chrome profile with the session companion. A profile selected during active research is queued and should complete through automatic sync; do not ask for a second activation click. Use noVNC for visible sign-in or challenge recovery.
  • Do not bypass X authentication, challenges, access controls, or rate limits.
  • For ChatGPT web, use the opt-in chatgpt Secure MCP Tunnel profile. Help the user create a workspace-associated tunnel and a runtime key restricted to Tunnels Read + Use, then ask them to enter both directly into npm run tunnel:secure:configure; its key prompt is hidden. Never capture, print, repeat, or commit either value. Use npm run tunnel:chatgpt only as an explicitly short-lived fallback, and never expose port 7345 directly.
  • Explain every visible step the user must perform and verify the final X identity without repeating its handle in public logs.

Changing the project

  • Keep the architecture tool-only unless a real user workflow requires UI.
  • Preserve one persistent browser, concurrency two, request spacing, durable jobs, account-bound provenance, and adaptive backoff.
  • Do not add posting, generic browser automation, CAPTCHA handling, proxy rotation, or enforcement-evasion features.
  • Use synthetic parser fixtures. Never commit captured posts, account handles, browser data, exports, registrations, tunnel identifiers, credentials, temporary URLs, or local paths. The only image exception is an explicitly allowlisted public showcase capture under docs/assets that has been manually checked against SECURITY.md.
  • Run npm run validate, npm audit --audit-level=moderate, docker compose config --quiet, and the plugin/skill validators before release.
  • Treat live tests as real account traffic. Use the smallest canary that proves the change and do not leave test monitors running.

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