LIVE
Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|OpenAI models are coming to Google Cloud21/10/26 · OpenAI|OpenAI details ChatGPT's new advertising infrastructure21/10/26 · OpenAI|Liquid AI unveils LFM3, a base for specialized multimodal models27/10/26 · Liquid AI|US Police Departments Use AI to Flag 'Suspicious' Driving Behavior27/10/26|AllenAI questions whether LLMs can judge research novelty27/10/26 · AllenAI|Open-Source AI Tools Aim to Speed Up Dementia Research27/10/26 · NVIDIA|Solid-state transformers eyed as a fix for AI's power grid bottleneck27/10/26|OpenAI expands Oracle partnership to bring business apps into ChatGPT27/10/26 · OpenAI|OpenAI publishes a page announcing 'GPT-6.1 Terra', with no technical details available26/10/26 · OpenAI|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|OpenAI models are coming to Google Cloud21/10/26 · OpenAI|OpenAI details ChatGPT's new advertising infrastructure21/10/26 · OpenAI|Liquid AI unveils LFM3, a base for specialized multimodal models27/10/26 · Liquid AI|US Police Departments Use AI to Flag 'Suspicious' Driving Behavior27/10/26|AllenAI questions whether LLMs can judge research novelty27/10/26 · AllenAI|Open-Source AI Tools Aim to Speed Up Dementia Research27/10/26 · NVIDIA|Solid-state transformers eyed as a fix for AI's power grid bottleneck27/10/26|OpenAI expands Oracle partnership to bring business apps into ChatGPT27/10/26 · OpenAI|OpenAI publishes a page announcing 'GPT-6.1 Terra', with no technical details available26/10/26 · OpenAI|
MCPRisingofficial publisher

GitHub MCP

MCP : Standardises the connection between an AI application and systems that expose tools or data · requires tout client MCP

Official GitHub MCP server: issues, PRs, code, and Actions accessible to the agent.

Essential as soon as the agent works on a GitHub repository; replaces the former reference server.

Overall score
8.0 /10
Security: one caveat, see the card
Utility
8.5
Reliability
10.0
Security
One caveat
Ease
5 min · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

Working with an AI agent on a GitHub project often means manually copying and pasting code, issues, or CI logs to give the assistant context. GitHub MCP Server removes that friction by connecting your AI client directly to the GitHub API, with no intermediary.

The server exposes tools that the agent can call in natural language: reading a file from a repository, creating a pull request, checking Dependabot alerts, monitoring a GitHub Actions workflow, or managing notifications. Two deployment modes are available: a remote server hosted by GitHub (accessible via OAuth or token), and a local version you run yourself via Docker.

This tool is aimed at developers who already use a compatible MCP client (VS Code 1.101+, Cursor, Claude Desktop, Windsurf, etc.) and want their AI agent to act directly on their repositories without leaving their working environment.

It is not suited for those looking for a tool with no technical setup: installation requires managing an access token or an OAuth flow, and the agent gains real write access to your repositories, which calls for careful attention to the permissions granted.

What it offers

  • ·Repository reading: browse code, search files, analyze commits, and understand project structure.
  • ·Issue and pull request management: create, update, triage, and comment directly from the agent.
  • ·GitHub Actions workflow monitoring: track runs, analyze build failures, and manage releases.
  • ·Security analysis: review Dependabot alerts and code scanning findings.
  • ·Team collaboration: access discussions, manage notifications, and analyze repository activity.
  • ·Remote server hosted by GitHub: connect without a local server via OAuth or token, for VS Code 1.101+ and other compatible clients.
  • ·Local mode via Docker: self-hosted deployment for GitHub Enterprise Server environments or restricted network setups.
  • ·GitHub Enterprise support: compatible with GitHub Enterprise Cloud (ghe.com) and GitHub Enterprise Server via a custom host parameter.

In practice

Price
Free
Commercial use
Commercial use: yes
Available on
claude-code, cursor, vscode
Host
tout client MCP
Licence
MIT

Security and data

Rating
Security: one caveat, see the card
Your data
Data with the vendor
Maintenance
Active
Last release
21 October 2026

Links and repository checked on 25 October 2026 · added on 1 April 2025

How to install GitHub MCP

1. Check that your MCP client is compatible (VS Code 1.101+, Cursor, Claude Desktop, Windsurf, etc.). 2. For remote mode, add the JSON configuration block to your client's MCP settings, choosing either OAuth (recommended on VS Code) or a Personal Access Token (PAT). 3. If using a PAT, create it at github.com/settings/personal-access-tokens/new granting only the necessary permissions, and store it in an environment variable rather than hardcoding it in the config file. 4. For local mode, install Docker, then use the one-click install button or manually add the Docker command to your client's configuration. 5. Enable Agent mode in your client (e.g., the Agent mode toggle in Copilot Chat on VS Code) so the server starts and becomes available. 6. For GitHub Enterprise Server, add the GITHUB_HOST parameter with your instance URL (prefixed with https://) to the Docker configuration.
docker run -i --rm -e GITHUB_PERSONAL_ACCESS_TOKEN ghcr.io/github/github-mcp-server

In the same category

Whole category Feed (memory, context) →