gerrit-mcp-server
A Model Context Protocol server that connects AI coding agents to Gerrit code review, enabling change search, diff review, and comment publishing without leaving the session.
README
gerrit-mcp-server
A Model Context Protocol server that connects your AI coding agent to Gerrit code review.
Ask your agent to find the changes waiting on you, read a diff, draft line comments, and publish a review — without leaving the session and without pasting change numbers back and forth.
It speaks MCP over stdio, so it works with any MCP client: Claude Code, Codex, Cursor, Zed, Continue, or your own. It is a single static binary with no runtime dependencies. You self-host it; nothing is sent anywhere except to the Gerrit host you configure.
Quick start
You need Node installed only so that npx can fetch and run the right binary for your machine. The
server itself is Go and has no Node runtime dependency.
1. Create a Gerrit auth token. In Gerrit, go to Settings → HTTP Credentials and generate one. See Authentication below if your Gerrit is older.
2. Add the server to your MCP client.
<details open> <summary><b>Claude Code</b></summary>
claude mcp add gerrit \
--env GERRIT_URL=https://gerrit.example.com \
--env GERRIT_USER=your-username \
--env GERRIT_TOKEN=your-token \
-- npx -y @gyeonghokim/gerrit-mcp-server
</details>
<details> <summary><b>Codex</b></summary>
Add this to ~/.codex/config.toml, or to .codex/config.toml for a single trusted project:
[mcp_servers.gerrit]
command = "npx"
args = ["-y", "@gyeonghokim/gerrit-mcp-server"]
# npx downloads the binary on first run, which can exceed the 10s default.
startup_timeout_sec = 60
[mcp_servers.gerrit.env]
GERRIT_URL = "https://gerrit.example.com"
GERRIT_USER = "your-username"
GERRIT_TOKEN = "your-token"
Or let the CLI write it for you:
codex mcp add gerrit \
--env GERRIT_URL=https://gerrit.example.com \
--env GERRIT_USER=your-username \
--env GERRIT_TOKEN=your-token \
-- npx -y @gyeonghokim/gerrit-mcp-server
</details>
<details> <summary><b>Cursor, Zed, and other clients</b></summary>
Add this to the client's MCP configuration file:
{
"mcpServers": {
"gerrit": {
"command": "npx",
"args": ["-y", "@gyeonghokim/gerrit-mcp-server"],
"env": {
"GERRIT_URL": "https://gerrit.example.com",
"GERRIT_USER": "your-username",
"GERRIT_TOKEN": "your-token"
}
}
}
}
</details>
3. Ask for something. "What changes am I reviewing?" or "Summarise the diff on change 12345."
Authentication
Gerrit authenticates REST clients with HTTP Basic using a token from your account settings, and
expects authenticated requests to be prefixed with /a/. This server handles the prefix for you.
Generate a token under Settings → HTTP Credentials. On Gerrit 3.13 and newer you can name the
token and give it a lifetime (90 days, 1 year, and so on) — worth doing, so the credential this
server holds is scoped and expires on its own.
Older Gerrit versions call the same thing an HTTP password. It still works: that endpoint is now
an alias that creates a token with the id legacy.
The token is sent in an Authorization header. It never appears in a command line, a log line, or
an error message.
Configuration
All configuration is environment variables, so it lives in your MCP client config and nowhere else.
| Variable | Required | Default | Description |
|---|---|---|---|
GERRIT_URL |
yes | — | Base URL of the Gerrit host, for example https://gerrit.example.com |
GERRIT_USER |
yes | — | Your Gerrit username |
GERRIT_TOKEN |
yes | — | Auth token from Settings → HTTP Credentials |
GERRIT_ALLOW_WRITE |
no | false |
Set to true to register the tools that modify Gerrit |
GERRIT_TIMEOUT |
no | 30s |
Per-request timeout |
GERRIT_LOG_LEVEL |
no | info |
debug, info, warn, or error. Logs go to stderr |
Available tools
Read-only tools are always registered. Write tools are off unless you set
GERRIT_ALLOW_WRITE=true, so an agent cannot abandon a change or post a review by accident.
Read
| Tool | Description |
|---|---|
query_changes |
Search changes with Gerrit query syntax (status:open owner:self) |
get_change_details |
Full summary of one change |
get_commit_message |
Commit message of the current patch set |
list_change_files |
Files touched by the latest patch set |
get_file_diff |
Diff for one file in a change |
list_change_comments |
Published comments on a change |
list_draft_comments |
Your unpublished draft comments |
changes_submitted_together |
Changes that would submit alongside this one |
suggest_reviewers |
Reviewer suggestions for a change |
get_bugs_from_cl |
Bug ids referenced in the commit message |
Write — requires GERRIT_ALLOW_WRITE=true
| Tool | Description |
|---|---|
post_review_comment |
Add a draft comment on a line, or reply in a thread |
publish_drafts |
Publish your draft comments as a review |
delete_draft_comment |
Delete one draft comment |
delete_draft_comments |
Delete every draft on a change |
add_reviewer |
Add a reviewer or CC |
set_topic |
Set or clear the topic |
set_ready_for_review |
Take a change out of WIP |
set_work_in_progress |
Mark a change WIP |
create_change |
Create a change |
abandon_change |
Abandon a change |
revert_change |
Revert a change |
revert_submission |
Revert a whole submission |
Security
- The token stays in a header. It is never passed as a process argument, so it cannot be read
out of
ps, and it is never written to logs. - Read-only by default. Destructive tools are not registered unless you opt in.
- Don't commit your MCP config. A project-level
.mcp.jsonwithGERRIT_TOKENin it is a credential in your repository. Keep it in your user-level client config, or gitignore it. - Use a dedicated token with a lifetime, so it can be revoked without touching your other credentials.
- Your Gerrit permissions still apply. This server cannot see or do anything your account cannot.
Supported Gerrit versions
Built against the Gerrit 3.14 REST API and tested against it. Expected to work with 3.12 and newer; older versions are missing some of the draft comment endpoints.
Other ways to install
npx is the easy path, but the binary stands alone.
# Go toolchain
go install github.com/GyeongHoKim/gerrit-mcp-server/cmd/gerrit-mcp-server@latest
Or download a binary for your platform from the
releases page and point your MCP
client's command at it directly. No Node required.
Development
mise install # toolchain, pinned in mise.toml
just setup # dependencies and git hooks
just ci # everything CI runs
just --list # all tasks
See AGENTS.md for architecture, conventions, and the Gerrit API details worth knowing before you touch the client.
License
Using this at work is fine. ELv2 places exactly three restrictions on you: you may not offer this software to third parties as a hosted or managed service, you may not circumvent license key functionality, and you may not strip the copyright notices. Running it, modifying it, forking it, and deploying it across your engineering organisation are all expressly permitted.
Note that ELv2 is source-available rather than OSI-approved open source. If your organisation screens dependencies by license, it may need to be allowlisted.
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。