15 published servers

Best Version Control MCP Servers

Version control servers expose repository context and collaboration workflows to an assistant. Some work with local repositories, while others connect to a hosting platform and its APIs. Potential tasks include inspecting changes, finding issues, or reviewing a pull request, but the actual tool set must be checked for each project. Separate the need to understand code history from the authority to publish changes or manage a repository.

For local access, verify which paths the server can read and whether it can change the working tree. For hosted access, use credentials restricted to the repositories and operations needed. Pay attention to private source code, issue attachments, and other material that may be returned to the model. Try a disposable branch or test repository before enabling write operations. Tool responses should preserve useful identifiers, such as commit references and pull request links, so that a human can inspect the evidence. Compare platform support, setup effort, and maintenance status, then choose an integration that complements your existing review and merge rules rather than bypassing them.

Compare Version Control servers

# / ServerWhy pick itStarsLast commitRuntimeRemote
A broad GitHub tool catalog for code and collaboration, with installation details explicitly unconfirmed.Not recordedPart of the modelcontextprotocol/servers monorepoOct 13, 2026nodeNo
idea-reality-mcp helps developers and AI coding agents check proposed products for existing alternatives before building. Its idea_check tool searches software and discussion sources, returning competition evidence, trend signals, and pivot suggestions.825Sep 18, 2026pythonNo
git-mcp-server gives AI agents tools to inspect Git repositories, stage and commit changes, manage branches, and synchronize remotes. Session directory context and wrap-up guidance support work across repositories and release preparation.242Sep 26, 2026nodeNo
github-projects-mcp connects GitHub Projects V2 workflows to Claude and other MCP clients. It serves users managing project boards, with tools for inspecting items, creating issues and drafts, changing field values, and removing project entries.32May 22, 2025Not specifiedNo
Perforce file changes, changelists, streams, and workspaces can be managed through Claude with mcp-perforce. It serves developers using P4 who want assistant-driven version control with project-specific configuration and structured command responses.29Dec 19, 2025nodeNo
Instagit gives coding agents source-based answers about Git repositories through its ask_repo tool. It supports architecture exploration, API integration research, debugging, security review, and version migration planning, with file and line citations.20Feb 13, 2026nodeNo
Primavera P6 schedule analysts can use p6xer-mcp-server to inspect XER files through MCP-compatible AI clients. Built on PyP6XER, it supports activity queries, critical-path review, resource analysis, and earned-value calculations.11Mar 25, 2026pythonNo
mcp-git-commit-generator helps developers prepare Conventional Commit messages through an MCP client. It exposes repository status and packages staged-change context with instructions for an LLM to draft the message.9Oct 12, 2026Not specifiedNo
WSC helps writers and developers inspect prose through AI assistants for style issues and research-cited patterns associated with AI-generated text. It reports contextual findings, returns cleaned text for repeated words, and supports local file checks.9Oct 14, 2026nodeNo
AtomGit collaboration tools let AI assistants find repositories, manage issues and assignments, and work with pull requests. The integration serves developers using MCP clients to inspect branches, organize issue labels, and participate in repository discussions.8Jul 17, 2025nodeNo
mcp-git-commit-aider lets AI assistants commit staged Git changes with an “(aider)” marker on the committer name. It serves developers who want to identify AI-associated commits while working with an assistant in their editor.8May 13, 2025nodeNo
mcp-optimizer brings PuLP and OR-Tools optimization tools to MCP clients for planning and resource-allocation work. It serves developers and planners modeling assignments, routes, schedules, portfolios, and production decisions.5Nov 5, 2025pythonNo
GitHub Actions workflows can be inspected and controlled through an AI assistant with mcp-server-github-actions. It serves developers investigating CI failures, retrying jobs, checking artifacts, or starting workflows from an MCP client.4Aug 10, 2026nodeNo
github-action-trigger-mcp gives developers MCP tools to inspect GitHub Actions, dispatch workflows with inputs, and retrieve recent releases. It also supports enabling auto-merge on a selected pull request.3Apr 17, 2026Not specifiedNo
Give agents GitHub-native repository context and issue, review, workflow, and security operations.33,713Oct 15, 2026goNo

Ordered by featured status, editorial order, then repository stars. Stars may cover an entire monorepo and are a dated snapshot, not a quality or security guarantee.

Frequently asked questions

Start with the exact tools your task needs, then compare runtime requirements, supported transports, permissions, and recent project activity. Test with nonsensitive data before connecting a production account. A high star count is a discovery signal, not a security audit.

The directory is free to browse. A server’s code license and any connected provider’s charges are separate matters. Check the upstream license, API pricing, and account requirements before installing.

Compatibility depends on the server’s transport and the client’s supported configuration. Local servers use stdio; Streamable HTTP is the current remote transport. MCP spec 2026-07-28 reclassified legacy HTTP+SSE as Deprecated. Some clients still accept it for older servers; migrate to Streamable HTTP. Follow the client guides and verify the current upstream documentation.