MCP server by pluginmd
lark-mcp-cli — Lark/Feishu × Claude (MCP)
English | Tiếng Việt | 中文版
Drive your Lark/Feishu workspace from Claude — read mail, summarize meetings, triage tasks, send messages, create docs, approve requests — in plain language, right inside Claude Desktop (Cowork) or claude.ai (web). No Claude Code, no copy-paste, no coding.
This is lark-cli (the official Lark/Feishu CLI — 200+ commands across 18 business domains) plus a built‑in MCP server and a Cowork skill kit, so any Claude surface can operate Lark out of the box.
🚀 Lark MCP · Install · 21 MCP Tools · Cowork Skills · Full Docs · Security
🚀 Lark MCP — use Lark from Claude Desktop & web
What's new in this build: a native MCP server (
lark-cli mcp serve) that turns the CLI into the hands of Claude inside Lark — plus 23 ready‑made Cowork workflow skills. Built for business users, not just developers.
| | |
| --- | --- |
| 🧰 21 curated MCP tools | IM, Mail, Calendar, Docs, Base, Contact, Task, Drive, Sheets, Meetings, OKR + a generic lark_api escape hatch → tool reference |
| 🖥️ Claude Desktop (Cowork) | Local stdio transport — one config block, restart, done → desktop guide |
| 🌐 claude.ai (web) | Streamable‑HTTP transport + Cloudflare Tunnel + built‑in bearer‑token gate → web guide |
| 🧠 23 Cowork skills | morning-brief, inbox-zero, meeting-prep, task-prioritizer, approval-triage… speak naturally, Claude runs the chain → skills |
| 🔐 Safe by default | Dry‑run on every write, mail saved as draft until confirmed, audit log, OS‑keychain credentials, bearer‑token + constant‑time check → security |
How it works
You ──"What's on my plate today?"──▶ Claude (Desktop / Web)
│ picks a tool + args
▼
lark-cli mcp serve (on your machine)
│ runs lark-cli <verb> with YOUR auth
▼
open.larksuite.com / open.feishu.cn
The MCP server spawns lark-cli subprocesses per tool call, reusing the existing auth, profile, and keychain — so your data only ever goes to Lark, never a third party (in stdio/local mode).
60‑second quick start (Claude Desktop)
# 1. Build & install the binary (adds the `mcp` command)
./scripts/setup-mcp.sh # installs to ~/bin/lark-cli
lark-cli mcp tools # should list 21 tools
# 2. Log in to Lark once (browser OAuth, token stored in OS keychain)
lark-cli auth login
# 3. Wire into Claude Desktop config, then restart Claude Desktop:
# ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"lark-cli": {
"command": "/absolute/path/to/lark-cli",
"args": ["mcp", "serve"],
"env": { "NO_COLOR": "1" }
}
}
}
Now ask Cowork: "List my calendar today" or "Find the contact named Alice". → Full step‑by‑step: docs/02. Web (claude.ai): docs/03.
Expose to the web (claude.ai) — securely
export LARK_MCP_BEARER_TOKEN=$(openssl rand -hex 32) # secret stays in env
lark-cli mcp serve --transport http --addr 127.0.0.1:3000 --audit-log ~/.lark-mcp-audit.ndjson
cloudflared tunnel --url http://127.0.0.1:3000 # → public HTTPS URL
The HTTP endpoint enforces Authorization: Bearer <token> on / and /mcp (constant‑time, /health stays open). Never expose the tunnel without the token. Add the URL (https://…/mcp) as a Custom Connector in claude.ai → secure web guide.
📚 MCP documentation (business‑user friendly)
| Doc | What | | --- | --- | | docs/README | Index + 3‑step start | | 01 — Overview & value | Why MCP, ROI | | 02 — Claude Desktop setup | stdio, step‑by‑step | | 03 — claude.ai web setup | HTTP + tunnel + bearer token | | 04 — Login & permissions | user/bot, scopes | | 05 — Cowork skills | 23 workflows | | 06 — MCP tools | 21 tools reference | | 07 — Security & privacy | bearer, audit, data flow | | 08 — Troubleshooting | incl. known issues | | 09 — Update & maintenance | upgrades | | MCP_QUICKSTART | all MCP hosts (Cursor, Zed, Cline…) |
Building or extending tools? See
cmd/mcp/README.mdfor the bridge architecture and the/mcp-addworkflow.
Why lark-cli?
Why lark-cli?
- Agent-Native Design — 24 structured Skills out of the box, compatible with popular AI tools — Agents can operate Lark with zero extra setup
- Wide Coverage — 18 business domains, 200+ curated commands, 26 AI Agent Skills
- AI-Friendly & Optimized — Every command is tested with real Agents, featuring concise parameters, smart defaults, and structured output to maximize Agent call success rates
- Open Source, Zero Barriers — MIT license, ready to use, just
npm install - Up and Running in 3 Minutes — One-click app creation, interactive login, from install to first API call in just 3 steps
- Secure & Controllable — Input injection protection, terminal output sanitization, OS-native keychain credential storage
- Three-Layer Architecture — Shortcuts (human & AI friendly) → API Commands (platform-synced) → Raw API (full coverage), choose the right granularity
Features
| Category | Capabilities |
| ------------- |-----------------------------------------------------------------------------------------------------------------------------------|
| 📅 Calendar | View, create and update events, invite attendees, find meeting rooms, RSVP to invitations, check free/busy & time suggestions |
| 💬 Messenger | Send/reply messages, create and manage group chats, view chat history & threads, search messages, download media |
| 📄 Docs | Create, read, update, and search documents, read/write media & whiteboards |
| 📁 Drive | Upload and download files, search docs & wiki, manage comments |
| 📝 Markdown | Create, fetch, patch, and overwrite Drive-native .md files |
| 📊 Base | Create and manage tables, fields, records, views, dashboards, workflows, forms, roles & permissions, data aggregation & analytics |
| 📈 Sheets | Create, read, write, append, find, and export spreadsheet data |
| 🖼️ Slides | Create and manage presentations, read presentation content, and add or remove slides |
| ✅ Tasks | Create, query, update, and complete tasks; manage task lists, subtasks, comments & reminders |
| 📚 Wiki | Create and manage knowledge spaces, nodes, and documents |
| 👤 Contact | Search users by name/email/phone, get user profiles |
| 📧 Mail | Browse, search, read emails, send, reply, forward, manage drafts, watch new mail |
| 🎥 Meetings | Search meeting records, query meeting minutes artifacts and recordings |
| 🕐 Attendance | Query personal attendance check-in records |
| ✍️ Approval | Query approval tasks, approve/reject/transfer tasks, cancel and CC instances |
| 🎯 OKR | Query, create, update OKRs; manage objective & key results, alignments, indicators and progress. |
| 📋 Project | Meegle — manage work items, schedules, and data via the standalone meegle-cli (install separately) |
| 🔗 Apps | Develop, deploy HTML, web pages and applications |
Installation & Quick Start
Requirements
Before you start, make sure you have:
- Node.js (
npm/npx) - Go
v1.23+ and Python 3 (only required for building from source)
Quick Start (Human Users)
Note for AI assistants: If you are an AI Agent helping the user with installation, jump directly to Quick Start (AI Agent), which contains all the steps you need to complete.
Install
Choose one of the following methods:
Option 1 — From npm (recommended):
npx @larksuite/cli@latest install
Option 2 — From source:
Requires Go v1.23+ and Python 3.
git clone https://github.com/larksuite/cli.git
cd cli
make install
# Install CLI SKILL (required)
npx skills add larksuite/cli -y -g
Configure & Use
# 1. Configure app credentials (one-time, interactive guided setup)
lark-cli config init
# 2. Log in (--recommend auto-selects commonly used scopes)
lark-cli auth login --recommend
# 3. Start using
lark-cli calendar +agenda
Quick Start (AI Agent)
The following steps are for AI Agents. Some steps require the user to complete actions in a browser.
Step 1 — Install
npx @larksuite/cli@latest install
Step 2 — Configure app credentials
Run this command in the background. It will output an authorization URL — extract it and send it to the user. The command exits automatically after the user completes the setup in the browser.
lark-cli config init --new
Step 3 — Login
Same as above: run in the background, extract the authorization URL and send it to the user.
lark-cli auth login --recommend
Step 4 — Verify
lark-cli auth status
Agent Skills
| Skill | Description |
| ------------------------------- |----------------------------------------------------------------------------------------------------------------|
| lark-shared | App config, auth login, identity switching, scope management, security rules (auto-loaded by all other skills) |
| lark-calendar | Calendar events (create/update), agenda view, free/busy queries, time suggestions, room finding, RSVP replies |
| lark-im | Send/reply messages, group chat management, message search, upload/download images & files, reactions |
| lark-doc | Create, read, update, search documents (Markdown-based) |
| lark-drive | Upload, download files, manage permissions & comments |
| lark-markdown | Create, fetch, patch, and overwrite Drive-native Markdown files |
| lark-sheets | Create, read, write, append, find, export spreadsheets |
| lark-slides | Create and manage presentations, read presentation content, and add or remove slides |
| lark-base | Tables, fields, records, views, dashboards, data aggregation & analytics |
| lark-task | Tasks, task lists, subtasks, reminders, member assignment |
| lark-mail | Browse, search, read emails, send, reply, forward, draft management, watch new mail |
| lark-contact | Search users by name/email/phone, get user profiles |
| lark-wiki | Knowledge spaces, nodes, documents |
| lark-event | Real-time event subscriptions (WebSocket), regex routing & agent-friendly format |
| lark-vc | Search meeting records, query meeting minutes (summary, todos, transcript) |
| lark-whiteboard | Whiteboard/chart DSL rendering |
| lark-minutes | Minutes metadata & AI artifacts (summary, todos, chapters); upload audio/video to create minutes, download media |
| lark-openapi-explorer | Explore underlying APIs from official docs |
| lark-skill-maker | Custom skill creation framework |
| lark-attendance | Query personal attendance check-in records |
| lark-approval | Query approval tasks, approve/reject/transfer tasks, cancel and CC instances |
| lark-workflow-meeting-summary | Workflow: meeting minutes aggregation & structured report |
| lark-workflow-standup-report | Workflow: agenda & todo summary |
| lark-okr | Query, create, update OKRs; manage objective & key results, alignments and indicators. |
Authentication
| Command | Description |
| ------------- | -------------------------------------------------------------- |
| auth login | OAuth login with interactive selection or CLI flags for scopes |
| auth logout | Sign out and remove stored credentials |
| auth status | Show current login status and granted scopes |
| auth check | Verify a specific scope (exit 0 = ok, 1 = missing) |
| auth scopes | List all available scopes for the app |
| auth list | List all authenticated users |
# Interactive login (TUI guides domain and permission level selection)
lark-cli auth login
# Filter by domain
lark-cli auth login --domain calendar,task
# Recommended auto-approval scopes
lark-cli auth login --recommend
# Exact scope
lark-cli auth login --scope "calendar:calendar:read"
# Agent mode: return verification URL immediately, non-blocking
lark-cli auth login --domain calendar --no-wait
# Resume polling later
lark-cli auth login --device-code <DEVICE_CODE>
# Identity switching: execute commands as user or bot
lark-cli calendar +agenda --as user
lark-cli im +messages-send --as bot --chat-id "oc_xxx" --text "Hello"
Three-Layer Command System
The CLI provides three levels of granularity, covering everything from quick operations to fully custom API calls:
1. Shortcuts
Prefixed with +, designed to be friendly for both humans and AI, with smart defaults, table output, and dry-run previews.
lark-cli calendar +agenda
lark-cli im +messages-send --chat-id "oc_xxx" --text "Hello"
lark-cli docs +create --api-version v2 --doc-format markdown --content $'<title>Weekly Report</title>\n# Progress\n- Completed feature X'
Run lark-cli <service> --help to see all shortcut commands.
2. API Commands
Auto-generated from Lark OAPI metadata, curated through evaluation and quality gates — 100+ commands mapped 1:1 to platform endpoints.
lark-cli calendar calendars list
lark-cli calendar events instance_view --params '{"calendar_id":"primary","start_time":"1700000000","end_time":"1700086400"}'
3. Raw API Calls
Call any Lark Open Platform endpoint directly, covering 2500+ APIs.
lark-cli api GET /open-apis/calendar/v4/calendars
lark-cli api POST /open-apis/im/v1/messages --params '{"receive_id_type":"chat_id"}' --data '{"receive_id":"oc_xxx","msg_type":"text","content":"{\"text\":\"Hello\"}"}'
Advanced Usage
Output Formats
--format json # Full JSON response (default)
--format pretty # Human-friendly formatted output
--format table # Readable table
--format ndjson # Newline-delimited JSON (for piping)
--format csv # Comma-separated values
Pagination
--page-all # Auto-paginate through all pages
--page-limit 5 # Max 5 pages
--page-delay 500 # 500ms between page requests
Dry Run
For commands that may have side effects, preview the request with --dry-run first:
lark-cli im +messages-send --chat-id oc_xxx --text "hello" --dry-run
Schema Introspection
Use schema to inspect any API method's parameters, request body, response structure, supported identities, and scopes:
lark-cli schema
lark-cli schema calendar.events.instance_view
lark-cli schema im.messages.delete
Security & Risk Warnings (Read Before Use)
This tool can be invoked by AI Agents to automate operations on the Lark/Feishu Open Platform, and carries inherent risks such as model hallucinations, unpredictable execution, and prompt injection. After you authorize Lark/Feishu permissions, the AI Agent will act under your user identity within the authorized scope, which may lead to high-risk consequences such as leakage of sensitive data or unauthorized operations. Please use with caution.
To reduce these risks, the tool enables default security protections at multiple layers. However, these risks still exist. We strongly recommend that you do not proactively modify any default security settings; once relevant restrictions are relaxed, the risks will increase significantly, and you will bear the consequences.
We recommend using the Lark/Feishu bot integrated with this tool as a private conversational assistant. Do not add it to group chats or allow other users to interact with it, to avoid abuse of permissions or data leakage.
Please fully understand all usage risks. By using this tool, you are deemed to voluntarily assume all related responsibilities.
Star History
Contributing
Community contributions are welcome! If you find a bug or have feature suggestions, please submit an Issue or Pull Request.
For major changes, we recommend discussing with us first via an Issue.
Before opening a PR, see AGENTS.md for the local build, test, and PR checklist used by contributors and AI agents.
License
This project is licensed under the MIT License. When running, it calls Lark/Feishu Open Platform APIs. To use these APIs, you must comply with the following agreements and privacy policies: