mcp-github-cli
About
a MCP server for github using the GH cli under the covers
Details
- Author
- michaelneale
- Downloads
- 314
- Categories
- Developer Tools
Jump to
- Get detailed authenticated user information via gh_get_me
- Execute GraphQL queries for repos, users, PRs, and issues
- Run custom GraphQL queries with gh_custom_graphql
- Search repositories using the REST API
- Create issues and pull requests via REST endpoints
- List branches, commits, and repository contents
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
mcp-github-cliCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Run with uvx mcp-github-cli from the command line. Prerequisites are installing the GitHub CLI (gh) and authenticating via gh auth login. Can be used with Goose or Claude. For a test run, execute uv run python main.py --test from the source directory.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"mcp-github-cli": {
"mcp-github-cli": {
"command": "uvx",
"args": [
"mcp-github-cli"
]
}
}
}
}
McpServers
{
"mcp-github-cli": {
"command": "uvx",
"args": [
"mcp-github-cli"
]
}
}
mcp-github-cli
MCP server that provides focused GraphQL and REST API tools for interacting with GitHub. This server offers a streamlined set of powerful tools that leverage GitHub's APIs to provide rich data and functionality.
Usage
uvx mcp-github-cli
Will run from PyPI. This can be used in goose or claude.
Prerequisites
1. Install the GitHub CLI (gh) if not already installed:
- macOS: brew install gh
- Linux: sudo apt install gh
- Windows: winget install GitHub.cli
2. Authenticate with GitHub:
gh auth login
Test
uv run python main.py --test
This will run a comprehensive test suite that verifies:
- Authentication status with GitHub
- User information retrieval (gh_get_me)
- GraphQL query functionality (viewer info)
- REST API repository search
- GraphQL repository information retrieval
- REST API branch listing
The test output provides a good demonstration of the capabilities of the different tools.
Usage from source
Running from CLI (Goose, or to try it)
uv --directory /path/to/mcp-github-cli run python main.py
Features
This MCP server focuses on providing powerful, focused tools that leverage GitHub's GraphQL and REST APIs:
User Information
- gh_get_me(): Get detailed information about the authenticated user
GraphQL Tools
GraphQL tools provide rich, nested data in a single request:
- gh_graphql_repo_info(owner_repo): Get comprehensive repository information
- gh_graphql_user_profile(username): Get detailed user profile data
- gh_graphql_pull_requests(owner_repo, limit, state): Get pull requests with reviewers and status
- gh_graphql_issues(owner_repo, limit, state, labels): Get issues with comments and labels
- gh_graphql_repo_contributors(owner_repo, limit): Get repository contributors and their commits
- gh_custom_graphql(query, variables): Execute a custom GraphQL query
REST API Tools
REST tools provide specific functionality for common operations:
- gh_rest_search_repos(query, limit): Search for repositories
- gh_rest_repo_contents(owner_repo, path, ref): Get repository contents
- gh_rest_create_issue(owner_repo, title, body, labels): Create an issue
- gh_rest_create_pr(owner_repo, title, body, head, base, draft): Create a pull request
- gh_rest_branches(owner_repo): List repository branches
- gh_rest_commits(owner_repo, branch, limit): List repository commits
- gh_rest_api(endpoint, method, data, query_params): Execute a custom REST API request
Examples
Here are examples of using these tools:
User Information
```python
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.





