Bitbucket DC

You enable AI assistants to interact with Bitbucket Data Center by running a Model Context Protocol (MCP) server. It provides a standardized interface to search code, browse files, manage pull requests, and explore repositories without leaving your IDE, making workflows faster and more integrated.

How to use

Install and run the MCP server in your IDE client to connect your AI assistant to your Bitbucket Data Center. You will configure a local MCP connection that the editor can load automatically when you open a project.

How to install

Prerequisites: you need Python and the uvx runtime available on your system.

1. Prepare your environment variables in your IDE configuration as shown in the code example.

{
  "mcpServers": {
    "bitbucket-dc": {
      "command": "uvx",
      "args": ["mcp-bitbucket-dc"],
      "env": {
        "BITBUCKET_HOST": "git.yourcompany.com",
        "BITBUCKET_API_TOKEN": "your-personal-access-token"
      }
    }
  }
}

Additional configuration and start

The server starts automatically when your IDE connects using the configured MCP server. Ensure you provide either the host or the full base URL for Bitbucket Data Center.

Security and tokens

Create a personal access token in Bitbucket Data Center with at least Repository Read permissions, and add it to your MCP configuration as shown. This token is used by the MCP server to access repository data.

Usage patterns

Ask your AI assistant to perform common tasks such as searching code, listing open pull requests, or navigating the repository structure. Examples include asking to search for a symbol across projects, list branches in a repository, or view PR details.

Troubleshooting

If the MCP server does not start when your IDE connects, verify the environment variables are correct and that the Bitbucket host and token are valid. Check for network access from your development machine to the Bitbucket Data Center instance.

Notes

This MCP server is built with Python using FastMCP and installs via the uvx runtime, so Node.js is not required.

Available tools

bitbucket_code_search

Search code across repositories using Lucene syntax to filter by extension, language, repository, project, path, and boolean operators.

bitbucket_browse

Browse directory trees to explore files and folders at a given path within a repository.

bitbucket_get_file_content

Retrieve raw file content with syntax highlighting for quick review.

bitbucket_list_files

Recursively list all file paths in a repository to map the entire codebase.

bitbucket_get_branches

List branches in a repository with optional filters.

bitbucket_get_tags

List tags in a repository with optional filters.

bitbucket_get_projects

List projects in Bitbucket Data Center with optional name/permission filters.

bitbucket_get_project

Get detailed information about a specific project.

bitbucket_get_repositories

List repositories within a project.

bitbucket_get_repository

Get repository details including clone URLs.

bitbucket_get_pull_requests

List pull requests with filters for state, direction, and text search.

bitbucket_get_pull_request

Get details of a specific pull request including reviewers.

bitbucket_get_pull_request_comments

Retrieve comments and activity on a pull request.

bitbucket_get_pull_request_changes

Get the files changed in a pull request.

bitbucket_get_pull_request_diff

Get the diff for a file in a pull request.

bitbucket_post_pull_request_comment

Post a comment on a pull request, either general or inline.

bitbucket_create_pull_request

Create a new pull request.

bitbucket_update_pull_request

Update PR title, description, or reviewers.

bitbucket_get_required_reviewers

Fetch the required reviewers for a targeted branch.

bitbucket_get_commits

List commits with optional path and reference range filters.