agentleFS
ClypmintClypmint/clypmint-broll/AGENTS.md

clypmint-broll

AGENTS.md4 starsChanged 17 days agoMITView on GitHub

This repository is the public, MIT side of the Clypmint B-Roll API: the Claude Code and Codex plugins, the MCP setup, the agent skill, the clypmint-broll CLI, the JavaScript and Python clients, and B-Roll Bench v1. The search engine runs at clypmint.com and is not in this repository. Its code lives in Clypmint's private repository, whose integrations/broll folder is a copy of this one. Change github.com/Clypmint/clypmint-broll first; the copy follows. These win over anything else in this file. 1. No…

  • 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 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 repository2 other files it gives its agents.

The file

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    No shell: Every child process gets an argument list.

    line 16
  2. Very specific:

    The key stays secret: Read it from CLYPMINT_API_KEY (or the plugin's secure store) and send it only to the API base URL.

    line 18
  3. Very specific:

    Everything here is public: No keys, internal hosts, customer data or internal numbers in code, comments or commit messages.

    line 21
  4. Distinctive:

    Use commit and pull-request titles that describe the change without private names, assistant chatter, or internal planning language.

    line 45

≈ “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

clypmint-broll: guide for coding agents

<!-- CLAUDE.md imports this file; Codex and Cursor read it directly. -->

This repository is the public, MIT side of the Clypmint B-Roll API: the Claude Code and Codex plugins, the MCP setup, the agent skill, the clypmint-broll CLI, the JavaScript and Python clients, and B-Roll Bench v1. The search engine runs at clypmint.com and is not in this repository. Its code lives in Clypmint's private repository, whose integrations/broll folder is a copy of this one. Change github.com/Clypmint/clypmint-broll first; the copy follows.

Rules

These win over anything else in this file.

  1. No dependencies. JavaScript uses Node built-ins (Node 18+; capture needs 22). Python uses the standard library (3.8+).
  2. No shell. Every child process gets an argument list. Put -- before a URL handed to another program. Text from a user or the API never becomes a flag.
  3. The key stays secret. Read it from CLYPMINT_API_KEY (or the plugin's secure store) and send it only to the API base URL. Never log it, print it or write it to disk.
  4. One JSON line. With --json, every CLI command prints exactly one JSON line, errors included.
  5. Everything here is public. No keys, internal hosts, customer data or internal numbers in code, comments or commit messages. Placeholders look like cm_live_....
  6. Tests never call the live API. They use a local fake server.
  7. One version. package.json, .claude-plugin/plugin.json, .codex-plugin/plugin.json, python/pyproject.toml and python/clypmint_broll/__init__.py carry the same version, and every pinned install line (#vX.Y.Z) names it. test/repo.test.mjs checks this.

Layout

PathWhat it is
src/client.mjsThe JavaScript client: search, plan, usage
src/cli.mjs, bin/The CLI
src/media.mjsCutting with ffmpeg and yt-dlp, and downloading
src/capture.mjsRecording a live site with the installed Chrome over the DevTools protocol
python/The Python client
skills/broll/SKILL.mdThe agent skill every plugin ships
.claude-plugin/, .mcp.jsonThe Claude Code plugin and its marketplace
.codex-plugin/, .codex-mcp.jsonThe Codex plugin (the key comes from CLYPMINT_API_KEY)
examples/Runnable examples, tested by test/examples.test.mjs
bench/v1/B-Roll Bench v1

Repository presentation

Use commit and pull-request titles that describe the change without private names, assistant chatter, or internal planning language. Preserve published history. Keep the README focused on installation, capabilities, verification, and support. Never describe a scan as proof of perfect security.

Tests

npm test
cd python && PYTHONPATH=. python3 -m unittest discover -s tests

Releasing

  1. Bump the version in the five files of rule 7, move every #vX.Y.Z pin to it, and add a CHANGELOG section.
  2. Open a pull request. After it merges, tag the merge commit vX.Y.Z and publish a GitHub release.
  3. Release tags are permanent: they cannot be moved or deleted. A mistake gets a new version.

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.