convez.ai
SkillIntegração & APIsIntermediário
xujie8867

Claude Code Lark Bridge

xujie8867/claude-code-lark-bridge

26 skills para integrar Claude Code com Lark/Feishu enterprise

Ver no GitHub
BomPluginScore · de 0 a 100

Coleção de 26 skills que conectam Claude Code ao Lark/Feishu, permitindo que agentes de IA executem operações empresariais completas. Abrange comunicação, documentos, dados, produtividade, reuniões e automação através de uma arquitetura baseada em CLI.

Instalação

$npx convez-ai xujie8867/claude-code-lark-bridge

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
1
Forks
0
Acompanhando
0
Saúde
85%
Último commit
há 3 dias
Versão
v1.0.0
Issues abertas
3
Contribuidores
2

Casos de uso

  • Automatizar envio e leitura de mensagens no Lark IM
  • Criar e gerenciar tabelas multi-dimensionais (Base)
  • Automação de workflows de aprovação e tarefas
  • Gerenciamento de calendários e agendamento de reuniões
  • Manipulação de documentos, planilhas e apresentações
  • Deploy de aplicações HTML em Miaoda

Veredito

Solução robusta e bem-documentada para integração profunda entre Claude Code e ecossistema Lark. Cobertura ampla de funcionalidades empresariais com arquitetura modular clara, ideal para automatizar operações em corporações que usam Feishu.

README

A tradução para o português deste README ainda está na fila.

Claude Code Lark Bridge


What is this?

Claude Code Lark Bridge is a collection of 26 custom Skills that give Claude Code (and any compatible AI agent) the ability to interact with Lark/Feishu — the enterprise collaboration platform used by millions.

Once installed, your AI agent can:

  • 📨 Send and read messages in Lark IM
  • 📊 Create and manage Base (multi-dimensional tables)
  • 📝 Read, write, and edit Docs, Sheets, and Slides
  • 📅 Schedule meetings and manage calendars
  • 📧 Send and read emails
  • ✅ Manage tasks, approvals, and OKRs
  • 🎥 Query video conference records and minutes
  • 🚀 Deploy HTML apps to Miaoda (飞书妙搭)
  • ...and 16 more capabilities

👉 See real-world usage examples →

Architecture

graph LR
    CC[Claude Code] -->|Skill System| BRIDGE[Lark Bridge Skills]
    BRIDGE -->|CLI| LARK[lark-cli]
    LARK -->|OpenAPI| FS[Lark/Feishu Platform]
    
    FS --> IM[IM Messages]
    FS --> BASE[Base Tables]
    FS --> DOC[Docs/Sheets/Slides]
    FS --> CAL[Calendar/VC]
    FS --> MAIL[Mail]
    FS --> TASK[Tasks/OKRs/Approvals]

Available Skills (26 Modules)

CategorySkillCapability
Communicationlark-imSend/receive messages, manage group chats
lark-mailDraft, send, reply, forward, search emails
lark-contactResolve contacts by name/email
Documentslark-docRead & edit Docs/Wiki
lark-sheetsCreate & manipulate spreadsheets
lark-slidesCreate & edit presentations
lark-markdownManage Markdown files
lark-whiteboardQuery & edit whiteboards
Data & Storagelark-baseMulti-dimensional tables, views, dashboards
lark-driveCloud storage: upload, download, manage files
lark-wikiKnowledge base management
Productivitylark-calendarSchedule, search events, book rooms
lark-taskCreate & track tasks, subtasks
lark-okrObjectives & key results
lark-approvalProcess approvals
lark-attendanceCheck attendance records
Meetingslark-vcQuery meeting history, minutes, transcripts
lark-vc-agentBot joins/leaves meetings, captures events
lark-minutesVoice-to-text meeting notes
Deploymentlark-appsDeploy HTML to Miaoda (public URLs)
Automationlark-eventReal-time event streaming/subscribing
lark-skill-makerCreate custom lark-cli Skills
lark-openapi-explorerDiscover native OpenAPI endpoints
Workflowslark-workflow-meeting-summaryStructured meeting reports
lark-workflow-standup-reportDaily standup summaries
Setuplark-sharedAuth, login, identity management

Quick Start

Prerequisites

  • Node.js ≥ 18
  • A Lark/Feishu account
  • Claude Code (or any Claude-powered IDE)

1. Install lark-cli

npm install -g @larksuite/cli

2. Clone & Deploy Skills

git clone https://github.com/xujie8867/claude-code-lark-bridge.git
cp -r claude-code-lark-bridge/skills/* ~/.claude/skills/

3. Initialize & Auth

lark-cli config init --new

Scan the QR code to authorize the Lark app.

4. Verify

lark-cli auth status
# Expected: Bot identity: ready

5. Try It

In Claude Code, just ask naturally:

> "Check my recent chat messages"
> "Create a project tracking spreadsheet"
> "Schedule a team meeting for Friday 3pm"
> "Deploy this HTML to Miaoda"

The AI automatically detects Lark-related requests and loads the appropriate Skill.

Identity Modes

ModeIdentityUse Case
Bot--as botApp-level operations (default, works out of the box)
User--as userPersonal account operations (requires lark-cli auth login)

Windows Setup

npm install -g @larksuite/cli
git clone https://github.com/xujie8867/claude-code-lark-bridge.git
xcopy claude-code-lark-bridge\skills\* %USERPROFILE%\.claude\skills\ /E /I
lark-cli config init --new

Real-World Use Cases

  • 🏢 Enterprise automation: Auto-generate weekly reports from meeting minutes
  • 📊 Data pipelines: Sync Lark Base data to external systems
  • 📧 Email triage: AI-powered inbox sorting and drafting
  • 🤖 Chatbot operations: Manage Lark bots and auto-reply workflows
  • 📅 Smart scheduling: Find optimal meeting times across teams
  • 🚀 Rapid deployment: Deploy internal tools to Miaoda in one command

Project Structure

claude-code-lark-bridge/
└── skills/
    ├── lark-im/          # Instant messaging
    │   ├── SKILL.md      # Skill definition & instructions
    │   └── assets/       # Icons, templates
    ├── lark-base/        # Multi-dimensional tables
    ├── lark-doc/         # Documents
    ├── lark-sheets/      # Spreadsheets
    ├── lark-calendar/    # Calendar & meetings
    ├── lark-drive/       # Cloud storage
    ├── lark-mail/        # Email
    │   ├── SKILL.md
    │   ├── references/   # Detailed operation guides
    │   └── assets/       # Email templates
    ├── lark-task/        # Task management
    ├── lark-approval/    # Approvals
    ├── lark-okr/         # Objectives & key results
    ├── lark-wiki/        # Knowledge base
    ├── lark-vc/          # Video conferencing
    ├── lark-vc-agent/    # VC bot agent
    ├── lark-minutes/     # Meeting minutes
    ├── lark-slides/      # Presentations
    ├── lark-markdown/    # Markdown files
    ├── lark-apps/        # Miaoda deployment
    ├── lark-attendance/  # Attendance
    ├── lark-event/       # Event streaming
    ├── lark-openapi-explorer/  # API discovery
    ├── lark-skill-maker/ # Custom skill creation
    ├── lark-whiteboard/  # Whiteboards
    ├── lark-contact/     # Contacts
    ├── lark-shared/      # Auth & setup
    ├── lark-workflow-meeting-summary/  # Meeting reports
    └── lark-workflow-standup-report/   # Standup reports

Ecosystem Impact

Lark/Feishu serves 100M+ users across Asia-Pacific enterprises. As AI agents become central to enterprise workflows, bridging them to existing collaboration platforms is critical infrastructure. This project fills that gap for the Claude Code ecosystem.

Roadmap

  • Add CI/CD for skill validation
  • English + Chinese bilingual skill descriptions
  • Web-based skill configurator
  • One-line install script
  • Video tutorials

Contributing

Contributions welcome! If you'd like to add a new skill or improve an existing one:

  1. Fork the repo
  2. Create a skill directory under skills/
  3. Add SKILL.md with the skill definition
  4. Submit a PR

License

MIT © 2026 xujie8867 (许海龙)


Documentation

One-Line Install

curl -fsSL https://raw.githubusercontent.com/xujie8867/claude-code-lark-bridge/master/scripts/install.sh | bash

Validation

./scripts/validate-skills.sh
# Expected: 26 skills | 0 errors | 0 warnings
Ver o original no GitHub

Bom para

Relacionados