LLM Markdown

Base Path: /v1/llm-markdown

Generates structured Markdown summaries of GitHub repositories, optimized for LLM consumption. Includes repository metadata, directory structure, and README content in a standardized format.

Request

Headers

• Content-Type: application/json

• Authorization: Bearer <your_token>

Body Parameters

Parameter	Type	Required	Description
repoUrl	string	Yes*	Full GitHub repository URL
repoOwner	string	Yes*	Repository owner/organization
repoName	string	Yes*	Repository name

* Either repoUrl OR both repoOwner and repoName must be provided

Example Request (using repoUrl)

{
  "repoUrl": "https://github.com/facebook/react"
}

Example Request (using owner/name)

{
  "repoOwner": "facebook",
  "repoName": "react"
}

Response

Success Response (200 OK)

Content-Type: text/markdown

# facebook/react

**Description:** A declarative, efficient, and flexible JavaScript library for building user interfaces.

**Stars:** 210,000+  
**Last Updated:** 2024-03-14  
**License:** MIT

[View on GitHub](https://github.com/facebook/react)

## Repository Structure

### Key Directories
- 📂 build - Production build outputs
- 📂 packages - Core packages and components
- 📂 fixtures - Test fixtures and examples
- 📂 scripts - Build and maintenance scripts

### Important Files
- 📄 package.json - Project configuration
- 📄 LICENSE - MIT License
- 📄 CONTRIBUTING.md - Contribution guidelines

### File Tree (First 20)

```text
.
├── build/
│   ├── dist/
│   └── node_modules/
├── packages/
│   ├── react/
│   ├── react-dom/
│   └── react-reconciler/
├── package.json
├── yarn.lock
└── README.md

README Content

[Repository README content here...]

### Error Responses

**Response (400 Bad Request):**

```json
{
  "status": 400,
  "data": null,
  "message": "Repository URL or owner/name combination required"
}

Response (401 Unauthorized):

{
  "status": 401,
  "data": null,
  "message": "Invalid or missing API key"
}

Response (404 Not Found):

{
  "status": 404,
  "data": null,
  "message": "Repository not found"
}

Usage Notes

• The generated Markdown is optimized for LLM context generation

• File tree is limited to 20 items for conciseness

• README content is truncated if exceeds 10,000 characters

• Repository metadata is cached for 1 hour

• Rate limits: 100 requests per hour per API key