bitbucket-mcp
Enables AI assistants to programmatically manage Bitbucket Cloud resources, including pull requests, repositories, and branches, automating code review workflows.
README
bitbucket-mcp
A Model Context Protocol (MCP) server that gives AI assistants (like Claude) programmatic access to Bitbucket Cloud. It enables reading and managing pull requests, browsing repositories and branches, and automating code review workflows from your MCP client.
It authenticates with Bitbucket using an app password tied to your Atlassian account. Multiple workspace connections can be configured, with an optional default repository per workspace.
What You Can Do
- "List all open PRs in my-repo targeting the main branch"
- "Create a pull request from feature/SCC-1234 to develop with a summary of the changes"
- "Show me what files changed in PR #42 and give me a line-by-line diff of the auth module"
- "Find the PR for branch feature/payments and merge it with a squash strategy"
- "Decline the stale PRs that have been open for more than 30 days"

Features
- Multi-workspace support: configure and switch between multiple Bitbucket Cloud workspaces
- Default repository: set a default repo per workspace so you don't need to specify it every call
- Web UI: browser-based management console for adding/editing workspace connections
- Protected branches: mark branches as protected to prevent accidental AI-triggered merges
- Branch autocomplete: the UI fetches and caches branch names for fast autocomplete
- Real-time activity log: live MCP tool execution log via SSE streamed to the Web UI
- Mock OAuth: built-in OAuth stub so MCP clients that require OAuth flows work out of the box
Tools
| Tool | Description |
|---|---|
list_workspaces |
List all configured Bitbucket workspace connections |
list_repos |
List repositories in a workspace |
list_branches |
List branches in a repository with optional name filter |
get_pull_requests |
List pull requests filtered by state, source, or destination branch |
get_pull_request |
Get details of a single pull request |
create_pull_request |
Open a new pull request |
merge_pull_request |
Merge a pull request (merge commit, squash, or fast-forward) |
decline_pull_request |
Decline a pull request |
get_pr_diff |
List all files changed in a PR with line counts |
get_pr_file_diff |
Get the full diff for a specific file in a PR |
Installation
Via npx (recommended)
npx @yunusemregul/bitbucket-mcp
Global install
npm install -g @yunusemregul/bitbucket-mcp
bitbucket-mcp
The server starts on http://localhost:18434 by default.
Options:
-p, --port Port to listen on (default: 18434)
-v, --version Print version
-h, --help Show help
Workspace configuration is stored in ~/.bitbucket-mcp/workspaces.json.
Configuration
Via Web UI
Open http://localhost:18434/ in your browser, click + Add Workspace, fill in the details (connection is tested automatically as you type), then click Save.

Workspace options
| Field | Type | Description |
|---|---|---|
name |
string | Display name for this connection |
workspaceSlug |
string | Bitbucket workspace slug (from the URL) |
username |
string | Atlassian account email |
token |
string | Bitbucket app password |
repoSlug |
string | Default repository slug (optional) |
protectedBranches |
string[] | Branches the AI is not allowed to merge into |
Tip: Set
protectedBranchesto["main", "master"]on production workspaces to block accidental merges.
Using with Claude
Claude Code (recommended)
claude mcp add --transport sse bitbucket-mcp http://localhost:18434/mcp/sse
Other MCP clients
{
"mcpServers": {
"bitbucket-mcp": {
"url": "http://localhost:18434/mcp/sse"
}
}
}
Project Structure
bitbucket-mcp/
├── server.js # Express app, MCP SSE endpoint, REST API
├── bitbucket.js # Bitbucket Cloud API client
├── storage.js # Workspace config persistence
├── tools/
│ ├── index.js # Tool registry
│ ├── context.js # Shared runtime state (sessions, logging)
│ └── *.js # One file per MCP tool
└── static/
├── index.html # Management console UI
├── app.js # UI logic
└── style.css # Styles
Security Notes
- Credentials are stored in plaintext in
~/.bitbucket-mcp/workspaces.json. Avoid exposing this file. - Use Bitbucket app passwords with the minimum required scopes rather than your main account password.
- Use
protectedBranchesto prevent the AI from merging into sensitive branches.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。