Skip to main content

Talk to your GitHub repos, issues, and project boards in plain English from your terminal

Project description

GitHub MCP Agent

Talk to your GitHub repos, issues, and project boards in plain English from your terminal.

GitHub MCP Agent demo

You > list open issues in my raytracer repo
You > set priority of issue #42 to urgent
You > what's in progress on the project board?
You > which issues are assigned to maryam?

Install

pip install github-mcp-agent

Or with uv:

uv tool install github-mcp-agent

Setup

github-agent setup

Interactive wizard:

  1. GitHub token (validated immediately)
  2. AI provider — choose one:
    • AWS Bedrock — existing profile, access keys, or SSO
    • Anthropic API — API key from console.anthropic.com
    • OpenAI — API key from platform.openai.com
    • Google Gemini — API key from aistudio.google.com
    • GitHub Copilot — browser OAuth flow, requires a Copilot subscription
    • Local (Ollama) — picks from your installed models, no API key needed
  3. Model selection (scrollable menu)
  4. Pulls the GitHub MCP Docker image

Config saved to ~/.config/github-mcp-agent/.env.


Usage

github-agent            # start the agent
github-agent setup      # full setup wizard
github-agent provider   # switch AI provider and model
github-agent model      # switch model within current provider
github-agent token      # update GitHub token
github-agent config     # edit config file in $EDITOR
github-agent prompt     # customize the system prompt in $EDITOR
github-agent -v         # start with verbose tool output

Requirements

Requirement Notes
Python 3.10+
Docker Must be running — used for the GitHub MCP server
GitHub token Scopes: repo, read:org, projectgithub.com/settings/tokens
Provider credentials See setup wizard

For AWS Bedrock: enable Claude model access in the Bedrock console under Model access.

For Ollama: install from ollama.com, run ollama serve. Models with good tool-calling support: qwen2.5, llama3.1, mistral.

For GitHub Copilot: requires an active Copilot subscription (student pack, individual, or business).


How it works

You (terminal)
    │
    ▼
github-agent (CLI)
    │
    ├── AI model (Bedrock / Anthropic / OpenAI / Gemini / Copilot / Ollama)
    │
    └── GitHub MCP Server (Docker)
            │
            └── GitHub API

Uses the Strands Agents SDK with the GitHub MCP server.


Troubleshooting

GITHUB_TOKEN not set — run github-agent setup

Docker not running — start Docker Desktop or sudo systemctl start docker

Bedrock access denied — check IAM permissions and model access in your region

Ollama model not found — run ollama list to confirm the model name, re-run github-agent model

No module named github_mcp_agent — reinstall with pip install github-mcp-agent or uv tool install github-mcp-agent

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

github_mcp_agent-0.2.1.tar.gz (801.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

github_mcp_agent-0.2.1-py3-none-any.whl (21.9 kB view details)

Uploaded Python 3

File details

Details for the file github_mcp_agent-0.2.1.tar.gz.

File metadata

  • Download URL: github_mcp_agent-0.2.1.tar.gz
  • Upload date:
  • Size: 801.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.10.6 {"installer":{"name":"uv","version":"0.10.6","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for github_mcp_agent-0.2.1.tar.gz
Algorithm Hash digest
SHA256 dc562e8c7ff0cf7497c976ff1b2289e228ba37b03c05e3ac82dd108f027bf85e
MD5 3ec967fafa77a1e74b2e5620a5232fb4
BLAKE2b-256 fb42233b480751d12f7119c83568f469361a9cc989ec93726dde503d38872975

See more details on using hashes here.

File details

Details for the file github_mcp_agent-0.2.1-py3-none-any.whl.

File metadata

  • Download URL: github_mcp_agent-0.2.1-py3-none-any.whl
  • Upload date:
  • Size: 21.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.10.6 {"installer":{"name":"uv","version":"0.10.6","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for github_mcp_agent-0.2.1-py3-none-any.whl
Algorithm Hash digest
SHA256 eec2732f936a6562e4d4ee37eed0a425230baacc6b8b7b7afa41eafcbd74371b
MD5 35b2ad8b3cbc200f67de809046eefa49
BLAKE2b-256 de492fdba7847349776c87169c67c376aeeef30c9cbfc71088d529014ad349e1

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page