# Manage Lark Meetings and Minutes

Meeting records and live interactions are spread across several Lark objects and identifiers. This skill routes lark-cli workflows while preserving identity and permissions.

## Install

```bash
npx skillstore add larksuite/lark-meeting
```

## Metadata

- Status: approved
- Slug: larksuite-lark-meeting
- Version: 1.0.0
- Author version: 1.0.0
- Skillstore revision: r1
- Version status: valid
- Tree hash: 4d05f716f0f3aada1a77cccb746be81a7b1732582c8999eae9711c90fc280618
- Author: larksuite
- GitHub username: larksuite
- License: MIT
- Repository: https://github.com/larksuite/cli/tree/b343e6763958db594bdb739a658122443c1c0e65/skills/lark-meeting
- Ref: 426435baad1227cbeafa615c19db2e9f3e76f9af
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: filesystem, network, external\_commands
- Quality score: 77
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/larksuite-lark-meeting
- Manifest: https://skillstore.pages.dev/api/skills/larksuite-lark-meeting/manifest

## Capabilities

- Search historical meetings and discover active meetings using user or application identity.
- Retrieve meeting details, participant snapshots, recordings, Notes, Minutes, and related document identifiers.
- Read summaries, action items, chapters, keywords, transcripts, and shared meeting documents.
- Create Minutes from uploaded media and edit titles, summaries, action items, keywords, or speakers.
- Join or leave meetings with an application bot and send visible text or reactions.
- Apply for Minutes access and manage collaborators when the user explicitly requests permission changes.

## Use Cases

- Review a completed meeting: Find the correct meeting, retrieve its transcript, and produce a focused summary with action items.
- Maintain Minutes content: Create Minutes from media, correct speakers, update summaries, and verify each requested change.
- Support a live meeting: Use an authorized application bot to join, read current events, send requested messages, and leave.

## Prompt Templates

### Find a meeting

```
Find my Lark meeting titled [title] from [date]. Show matching meetings with times and organizers before selecting one.
```

### Summarize a transcript

```
Open the transcript for meeting [meeting ID]. Summarize decisions, unresolved questions, owners, and due dates from the original discussion.
```

### Create and verify Minutes

```
Create Lark Minutes from [local media path]. Return the link, wait until ready, then extract the transcript and verify completion.
```

### Coordinate a live meeting bot

```
Join meeting [meeting number] with the application bot. Read current events, send [message] after confirming content, then leave only when requested.
```

## Limitations

- Requires an installed and authenticated lark-cli with suitable Lark scopes and tenant access.
- Results remain limited by the selected identity, resource permissions, application installation, and data-access rules.
- Meeting scheduling, availability, and room management require the separate lark-calendar skill.
- Some transcript operations require user identity, and media downloads create local files.

## Best Practices

- Confirm the unique meeting, Minutes, or Note identifier before reading or changing content.
- Preserve the identity that produced each identifier and explain any required identity change.
- Require explicit user intent for visible messages, bot participation, edits, uploads, downloads, and permission changes.

## Anti Patterns

- Do not treat meeting numbers, meeting IDs, Note IDs, document tokens, and Minutes tokens as interchangeable.
- Do not switch identities to bypass permissions or repeat failed writes automatically.
- Do not claim asynchronous Minutes artifacts are ready until a follow-up check confirms completion.

## Security Audit

- Audited at: 2026-08-21T08:34:08.45\+00:00
- Summary: All 232 static findings are false positives caused by Markdown syntax, placeholder identifiers, readable Chinese text, and example response URLs. No prompt injection, obfuscated payload, path traversal, host reconnaissance, credential access, or hardcoded external endpoint was found.

## Stats

- Views: 0
- Downloads: 0
- Favorites: 0
- Popularity score: 0