agentsmd-init

KO | EN

gitlite — search

#gemini #cloud-native #llm #tailwindcss #csharp #kubernetes #docker #markdown #monitoring #security

(★ 20)

Scaffold a high-quality AGENTS.md from a repository scan.

File Explorer

ZIP Download

- - __init__.py
  - cli.py
  - detect.py
  - render.py
- test_render.py
.gitignore
LICENSE
pyproject.toml
README.md

# Use via CDN

jsDelivr

jsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.

Version

File

Link

Example

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

**Scaffold a high-quality `AGENTS.md` from a repository scan.**

`AGENTS.md` is becoming the standard way to give AI coding agents the context they
need: how to set up, build, test, and behave inside a repository. Writing one by
hand from scratch is tedious and easy to get wrong. `agentsmd-init` scans your
project, detects the stack and common commands, and generates a clean, structured
`AGENTS.md` you can commit and refine.

Part of the AAIF (Agentic AI Innovation Foundation) open-source toolset for the
agentic AI ecosystem.

## Why

- Agents work far better with an explicit `AGENTS.md` than by guessing.
- A consistent structure (setup, build/test, layout, safety, definition of done)
  saves reviewers time and reduces agent mistakes.
- Zero dependencies, standard library only. Runs anywhere Python runs.

## Install

```bash
pip install -e .
```

Requires Python 3.9 or newer. No third-party dependencies.

## Usage

Generate an `AGENTS.md` in the current repository:

```bash
agentsmd-init
```

Preview without writing a file:

```bash
agentsmd-init --stdout
```

Scan a different directory and overwrite an existing file:

```bash
agentsmd-init /path/to/repo --force
```

## What it detects

| Marker file        | Stack            | Commands generated              |
|--------------------|------------------|---------------------------------|
| `pyproject.toml`   | Python           | install, test, lint             |
| `requirements.txt` | Python           | install, test, lint             |
| `package.json`     | Node.js          | install, build, test, lint      |
| `go.mod`           | Go               | install, build, test            |
| `Cargo.toml`       | Rust             | install, build, test            |
| `pom.xml`          | Java (Maven)     | install, build, test            |

For Node projects it reads the real `scripts` block in `package.json` and prefers
those commands. Detected placeholders are marked `TODO` so nothing is silently
wrong.

## Generated structure

The output includes:

- **Project** summary and detected stack
- **Setup** commands
- **Build, test, and lint** table
- **Project layout** snapshot
- **Conventions** for agents to follow
- **Safety and boundaries** to prevent destructive actions
- **Definition of done**

Treat the result as a strong first draft. Fill in the project overview and adjust
any command the scanner could not infer.

## Development

```bash
pip install -e .
pytest
```

## License

MIT. See [LICENSE](LICENSE).

agentsmd-init

Owner Repos

File Explorer

# Use via CDN

Was this content helpful?

Recent Feedback

agentsmd-init

Owner Repos

File Explorer

# Use via CDN

Was this content helpful?

Recent Feedback

Download README