lark-meeting

Query Feishu video meetings, minutes, and AI notes via lark-cli commands.

8|2|Updated Apr 29, 2026
One-click install
npx skills add https://github.com/luojiaaoo/xpeech --skill lark-meeting-luojiaaoo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: lark-meeting
Source: https://github.com/luojiaaoo/xpeech/tree/main/xpeech/agent/skills/buildin/lark-skills/lark-meeting
Command: npx skills add https://github.com/luojiaaoo/xpeech --skill lark-meeting-luojiaaoo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? It lets an agent locate Feishu (Lark) video meetings and their artifacts—smart notes, transcripts, and Minutes recordings—and answer questions or perform actions on them without manually navigating the Feishu UI. ## Core Features & Use Cases - Meeting & artifact lookup: Search historical meetings by keyword, time, organizer, or participant; resolve meeting_id, meeting_no, note_id, and minute_token to fetch details, participants, and recordings. - Minutes management: Search, upload local audio/video to generate Minutes, read summaries/todos/chapters/transcripts, download media, and edit titles, summaries, todos, speakers, or keywords. - Live meeting interaction: Discover active meetings, read real-time events (speech, chat, shared docs), send in-meeting messages or reactions, and have the app bot join or leave meetings. - Use Case: Ask "summarize what was discussed in this morning's meeting" and the agent finds the meeting, pulls the smart note or Minutes transcript, and produces an independent recap. ## Quick Start Ask the agent to find today's finished Feishu meetings and summarize the key decisions from the meeting notes.

Frequently Asked Questions about lark-meeting

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search Feishu meeting records with lark-cli?▼

Use lark-cli vc +search with at least one filter such as --query, --start/--end time range, --organizer-ids, or --participant-ids. It only searches ended meetings with user identity and requires the vc:meeting.search:read scope.

How do I get the transcript of a Feishu meeting?▼

First resolve the meeting to its note_id via vc +detail or minute_token via vc +recording. Then use note +detail or note +transcript for smart notes, or minutes +detail --transcript for Minutes, keeping the same identity that obtained the identifier.

Can I upload a local audio or video file to Feishu Minutes?▼

Yes, but not directly. Upload the file to Feishu Drive first with drive +upload to get a file_token, then call minutes +upload --file-token. Files must be under 6 hours and 6 GB in supported audio or video formats.

What is the difference between meeting_id, note_id, and minute_token in Feishu?▼

meeting_id identifies the actual video meeting, note_id identifies the AI smart note from the AI summary pipeline, and minute_token identifies the Minutes recording artifact. They come from independent pipelines and cannot be substituted for one another.

Why does reading meeting events fail with bot is not in meeting?▼

The app identity can only read events for meetings the bot actually joined, or within a 5-minute grace window after the meeting ends if the bot attended. Use user identity for meetings discovered via user identity, or have the bot join first with vc +meeting-join.

Can the Feishu bot join a meeting automatically?▼

Only when the user explicitly requests it. The bot joins with vc +meeting-join --as bot --meeting-number using the 9-digit meeting number, and the join is visible to other participants, so it is treated as a write operation requiring clear intent.