Bitbucket Server MCP

Bitbucket Server MCP

An MCP server for Atlassian Bitbucket Server / Data Center, providing 66 tools for managing projects, repositories, branches, files, commits, pull requests, and code search with opt-in deletion operations.

Category
访问服务器

README

Bitbucket Server MCP

CI PyPI version Python License: MIT

An MCP (Model Context Protocol) server for Atlassian Bitbucket Server / Data Center (Enterprise). Provides 66 tools for reading and writing projects, repositories, branches, files, commits, pull requests, and code search — with opt-in gated deletion operations (disabled by default, see SECURITY.md.

Installation

From PyPI (recommended)

pip install bitbucket-server-mcp

Or with uv:

uv pip install bitbucket-server-mcp

From source

git clone https://github.com/ManpreetShuann/bitbucket-server-mcp.git
cd bitbucket-server-mcp
uv sync

Via Docker (no local Python required)

cd bitbucket-server-mcp
docker build -t bitbucket-server-mcp .

Prerequisites

  • Python 3.10+
  • Bitbucket Server 7.x+ with HTTP access tokens enabled

Configuration

Set two required environment variables (plus optional ones):

Variable Required Description
BITBUCKET_URL Yes Base URL of your Bitbucket Server (e.g., https://bitbucket.yourcompany.com). Must use HTTPS.
BITBUCKET_TOKEN Yes HTTP access token (create in Bitbucket > User Settings > HTTP Access Tokens)
BITBUCKET_LOG_LEVEL No Log level for stderr output: DEBUG, INFO, WARNING, ERROR (default: INFO)
BITBUCKET_ALLOW_DANGEROUS_DELETE No Set to 1 to enable Tier-1 delete tools (branch, tag, PR, comment, task, attachment)
BITBUCKET_ALLOW_DESTRUCTIVE_DELETE No Set to 1 to enable Tier-2 delete tools (project, repository). Requires BITBUCKET_ALLOW_DANGEROUS_DELETE=1

Integration

This server runs locally via stdio rather than as a hosted HTTP service. Running locally keeps your Bitbucket access token on your own machine, avoids exposing an authenticated API endpoint over the network, and removes the need to manage server infrastructure or TLS certificates.

Claude Code

Add to ~/.claude.json or project .claude/settings.json:

<details> <summary>With pip (recommended)</summary>

pip install bitbucket-server-mcp
{
  "mcpServers": {
    "bitbucket": {
      "command": "bitbucket-server-mcp",
      "env": {
        "BITBUCKET_URL": "https://bitbucket.yourcompany.com",
        "BITBUCKET_TOKEN": "your-token-here"
      }
    }
  }
}

</details>

<details> <summary>With Docker</summary>

{
  "mcpServers": {
    "bitbucket": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "-e", "BITBUCKET_URL", "-e", "BITBUCKET_TOKEN", "bitbucket-server-mcp"],
      "env": {
        "BITBUCKET_URL": "https://bitbucket.yourcompany.com",
        "BITBUCKET_TOKEN": "your-token-here"
      }
    }
  }
}

</details>

<details> <summary>With uv (from source, for development)</summary>

git clone https://github.com/ManpreetShuann/bitbucket-server-mcp.git
cd bitbucket-server-mcp
uv sync
{
  "mcpServers": {
    "bitbucket": {
      "command": "uv",
      "args": ["run", "--directory", "/path/to/bitbucket-server-mcp", "bitbucket-server-mcp"],
      "env": {
        "BITBUCKET_URL": "https://bitbucket.yourcompany.com",
        "BITBUCKET_TOKEN": "your-token-here"
      }
    }
  }
}

</details>

GitHub Copilot (VS Code)

Add to .vscode/mcp.json in your workspace:

<details> <summary>With pip (recommended)</summary>

pip install bitbucket-server-mcp
{
  "servers": {
    "bitbucket": {
      "type": "stdio",
      "command": "bitbucket-server-mcp",
      "env": {
        "BITBUCKET_URL": "https://bitbucket.yourcompany.com",
        "BITBUCKET_TOKEN": "${input:bitbucket-token}"
      }
    }
  },
  "inputs": [
    {
      "type": "promptString",
      "id": "bitbucket-token",
      "description": "Bitbucket Server HTTP access token",
      "password": true
    }
  ]
}

</details>

<details> <summary>With Docker</summary>

{
  "servers": {
    "bitbucket": {
      "type": "stdio",
      "command": "docker",
      "args": ["run", "--rm", "-i", "-e", "BITBUCKET_URL", "-e", "BITBUCKET_TOKEN", "bitbucket-server-mcp"],
      "env": {
        "BITBUCKET_URL": "https://bitbucket.yourcompany.com",
        "BITBUCKET_TOKEN": "${input:bitbucket-token}"
      }
    }
  },
  "inputs": [
    {
      "type": "promptString",
      "id": "bitbucket-token",
      "description": "Bitbucket Server HTTP access token",
      "password": true
    }
  ]
}

</details>

<details> <summary>With uv (from source, for development)</summary>

git clone https://github.com/ManpreetShuann/bitbucket-server-mcp.git
cd bitbucket-server-mcp
uv sync
{
  "servers": {
    "bitbucket": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "--directory", "/path/to/bitbucket-server-mcp", "bitbucket-server-mcp"],
      "env": {
        "BITBUCKET_URL": "https://bitbucket.yourcompany.com",
        "BITBUCKET_TOKEN": "${input:bitbucket-token}"
      }
    }
  },
  "inputs": [
    {
      "type": "promptString",
      "id": "bitbucket-token",
      "description": "Bitbucket Server HTTP access token",
      "password": true
    }
  ]
}

</details>

OpenAI Codex

Add to ~/.codex/config.toml:

<details> <summary>With pip (recommended)</summary>

pip install bitbucket-server-mcp
[mcp_servers.bitbucket]
command = "bitbucket-server-mcp"
args = []

[mcp_servers.bitbucket.env]
BITBUCKET_URL = "https://bitbucket.yourcompany.com"
BITBUCKET_TOKEN = "your-token-here"

Or via CLI:

codex mcp add bitbucket \
  --env BITBUCKET_URL=https://bitbucket.yourcompany.com \
  --env BITBUCKET_TOKEN=your-token-here \
  -- bitbucket-server-mcp

</details>

<details> <summary>With Docker</summary>

[mcp_servers.bitbucket]
command = "docker"
args = ["run", "--rm", "-i", "-e", "BITBUCKET_URL", "-e", "BITBUCKET_TOKEN", "bitbucket-server-mcp"]

[mcp_servers.bitbucket.env]
BITBUCKET_URL = "https://bitbucket.yourcompany.com"
BITBUCKET_TOKEN = "your-token-here"

Or via CLI:

codex mcp add bitbucket \
  --env BITBUCKET_URL=https://bitbucket.yourcompany.com \
  --env BITBUCKET_TOKEN=your-token-here \
  -- docker run --rm -i -e BITBUCKET_URL -e BITBUCKET_TOKEN bitbucket-server-mcp

</details>

<details> <summary>With uv (from source, for development)</summary>

git clone https://github.com/ManpreetShuann/bitbucket-server-mcp.git
cd bitbucket-server-mcp
uv sync
[mcp_servers.bitbucket]
command = "uv"
args = ["run", "--directory", "/path/to/bitbucket-server-mcp", "bitbucket-server-mcp"]

[mcp_servers.bitbucket.env]
BITBUCKET_URL = "https://bitbucket.yourcompany.com"
BITBUCKET_TOKEN = "your-token-here"

Or via CLI:

codex mcp add bitbucket \
  --env BITBUCKET_URL=https://bitbucket.yourcompany.com \
  --env BITBUCKET_TOKEN=your-token-here \
  -- uv run --directory /path/to/bitbucket-server-mcp bitbucket-server-mcp

</details>

Tools (66 total)

Projects

Tool Description
list_projects List all projects (paginated)
get_project Get project details by key

Repositories

Tool Description
list_repositories List repos in a project (paginated)
get_repository Get repo details
create_repository Create a new repo

Branches & Tags

Tool Description
list_branches List branches (paginated, filterable)
get_default_branch Get default branch
create_branch Create a branch from a start point
list_tags List tags (paginated, filterable)

Files & Content

Tool Description
browse_files Browse directory/file tree at a path and revision
get_file_content Get raw file content
list_files List file paths in a directory

Commits

Tool Description
list_commits List commits (paginated, filterable by branch/path)
get_commit Get commit details
get_commit_diff Get diff for a commit
get_commit_changes Get changed files for a commit

Pull Requests

Tool Description
list_pull_requests List PRs with state/direction/draft/title/participant filters (paginated)
get_pull_request Get PR details
create_pull_request Create a PR with reviewers (supports draft mode)
update_pull_request Update PR title/description/reviewers/draft status
create_draft_pull_request Create a PR in draft mode
publish_draft_pull_request Publish a draft PR (mark as ready for review)
convert_to_draft Convert an open PR back to draft
can_merge_pull_request Check merge readiness (canMerge, conflicts, vetoes)
merge_pull_request Merge a PR with optional strategy
decline_pull_request Decline a PR
reopen_pull_request Reopen a declined PR
approve_pull_request Approve a PR
unapprove_pull_request Remove your approval
request_changes_pull_request Request changes on a PR
remove_change_request_pull_request Remove your change request
list_pull_request_participants List reviewers with roles and statuses
watch_pull_request Subscribe as a watcher
unwatch_pull_request Unsubscribe from watching
get_commit_message_suggestion Get suggested commit message for merge
get_pull_request_diff Get PR diff
get_pull_request_diff_stat Get per-file change list (added/modified/deleted)
list_pull_request_commits List commits in a PR
get_pull_request_activities Get PR activity feed
list_pull_request_comments List comments on a PR
get_pull_request_comment Get a specific comment
add_pull_request_comment Add a comment (general, inline, reply, or blocker task)
update_pull_request_comment Edit a comment
resolve_pull_request_comment Resolve a comment thread
reopen_pull_request_comment Reopen a resolved thread
list_pull_request_tasks List PR tasks
create_pull_request_task Create a task (optionally linked to a comment)
get_pull_request_task Get a specific task
update_pull_request_task Update task content or state

Dashboard

Tool Description
list_dashboard_pull_requests List PRs visible to the authenticated user across all repos (paginated)
list_inbox_pull_requests List PRs in the user's inbox needing review action (paginated)

Search

Tool Description
search_code Search code across repos (requires Elasticsearch)
find_file Find files by name or path pattern with wildcards

Users

Tool Description
find_user Search users by name, username, or email

Attachments

Tool Description
get_attachment Download an attachment by ID
get_attachment_metadata Get attachment metadata
save_attachment_metadata Create or update attachment metadata

Dangerous Delete (Tier-1, requires BITBUCKET_ALLOW_DANGEROUS_DELETE=1)

Tool Description
delete_branch Delete a branch (irreversible)
delete_tag Delete a tag (irreversible)
delete_pull_request Delete a PR and all its contents
delete_pull_request_comment Delete a PR comment
delete_pull_request_task Delete a PR task
delete_attachment Delete an attachment
delete_attachment_metadata Delete attachment metadata

Destructive Delete (Tier-2, requires both BITBUCKET_ALLOW_DANGEROUS_DELETE=1 and BITBUCKET_ALLOW_DESTRUCTIVE_DELETE=1)

Tool Description
delete_project Delete a project and all its repositories (irreversible)
delete_repository Delete a repository and all its contents (irreversible)

Development

git clone https://github.com/ManpreetShuann/bitbucket-server-mcp.git
cd bitbucket-server-mcp
uv sync                # Install all dependencies (including dev)
uv run pytest -v       # Run tests
uv run ruff check src/ tests/   # Lint
uv run ruff format src/ tests/  # Format

See CONTRIBUTING.md for full development guidelines.

Pagination

All list tools accept start (default 0) and limit (default 25) parameters. Responses include isLastPage and nextPageStart for fetching subsequent pages.

Security

See SECURITY.md for our security policy and vulnerability reporting instructions.

License

MIT

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选