vc (v1)

vc (v1) Skill for Happycapy Skills Platform

What Is This

The vc (v1) skill (Skill ID: lark-vc) is an integration for the Happycapy Skills platform that enables advanced interaction with Feishu (Lark Suite) video meetings. It allows users to query historical meeting records, search for past meeting details, and retrieve structured meeting artifacts such as AI-generated summaries, to-dos, section breakdowns, and complete verbatim transcripts. The skill leverages the lark-cli command-line tool to interact with Feishu’s backend, providing automation and structured data extraction for meeting content.

This skill is essential for teams using Feishu video conferencing, offering streamlined access to meeting outputs without manual retrieval or document navigation. The core focus is on post-meeting data - upcoming and scheduled meetings are outside the scope and should be managed by the separate lark-calendar skill.

Why Use It

Efficient meeting management is a cornerstone of modern collaboration. The vc (v1) skill addresses several pain points in meeting follow-up:

• Automated Record Retrieval: Obtain comprehensive lists of past meetings, filtered by keywords, time ranges, participants, organizers, or meeting rooms.
• Structured Artifact Access: Access AI-generated meeting summaries, to-dos, section lists, and verbatim transcripts for knowledge sharing and compliance.
• Time Savings: Avoid manual searching and downloading of meeting assets, especially for teams with frequent or large meetings.
• Consistent Documentation: Ensure all meeting artifacts are collected and stored in standardized directories for easy future reference.
• Integration: As part of the Happycapy Skills ecosystem, it integrates with other tools in your workflow, supporting automation and batch operations.

This skill is ideal for organizations that rely heavily on Feishu video meetings and require a systematic approach to post-meeting knowledge management.

How to Use It

Prerequisites

• The lark-cli tool must be installed and available in your environment.
• Proper authentication and permission setup are required. You must first use the Read tool to read ../lark-shared/SKILL.md for authentication and permissions handling.

Core Commands and Usage

1. Search for Past Meetings

   You can search for already ended meetings using various filters:

   bash

   lark-cli vc +list --keyword "project review" --start "2024-06-01" --end "2024-06-30"

   Filters include keywords, time ranges, organizers, participants, and meeting rooms. Pagination must be handled to ensure all results are retrieved.

2. Retrieve Meeting Artifacts

   After identifying the meeting of interest (using its meeting_id), you can fetch its structured artifacts.

   • Fetch Meeting Minutes/Notes (MainDoc and VerbatimDoc):

     By default, you only need to provide links to the main note and verbatim transcript, not the actual content. If the user requests specific content (summary, to-dos, sections), you can fetch them as follows:

     bash

     # Fetch the content of the structured meeting note
     lark-cli docs +fetch --doc <note_doc_token>

   • Extract and Download Whiteboard Cover (if available):

     The meeting note may include a <whiteboard token="xxx"/> tag in the returned markdown. The first such tag represents the cover image (AI-generated summary visualization).

     bash

     # Download the whiteboard cover image
     lark-cli docs +media-download --type whiteboard --token <whiteboard_token> --output ./artifact-<title>/cover

     Not all meeting notes have a whiteboard cover. If none exists, this step can be skipped.

   • Directory Structure:

     All artifacts for a single meeting (cover image, verbatim transcript, notes, etc.) must be stored together under ./artifact-<title>/.

Example Workflow

1. Search for all meetings from last week organized by "Alice":

   bash

   lark-cli vc +list --organizer "Alice" --start "2024-06-10" --end "2024-06-16"

2. For a selected meeting, fetch the note document:

   bash

   lark-cli docs +fetch --doc <note_doc_token>

3. Parse the markdown for the first <whiteboard token="..."/> tag, then download the cover:

   bash

   lark-cli docs +media-download --type whiteboard --token <whiteboard_token> --output ./artifact-<title>/cover

When to Use It

• After Meetings Have Ended: This skill is explicitly for accessing and managing artifacts from completed meetings. For future or scheduled meetings, use the lark-calendar skill.
• When You Need Structured Outputs: Use vc (v1) when you need AI-generated summaries, to-dos, or verbatim transcripts for compliance, sharing, or knowledge management.
• For Batch Retrieval or Filtering: When searching across many meetings with specific filters (by time, person, or room), vc (v1) provides robust querying.
• For Artifact Download Automation: When you want to automatically download and organize all post-meeting assets into a single directory for each meeting.

Important Notes

• Authentication and Permissions: Always ensure you read and follow the guidance in ../lark-shared/SKILL.md before using this skill. Improper permissions will prevent access to meeting data.
• Ended Meetings Only: This skill does not support future or ongoing meetings. Attempting to use it for those cases will result in errors or empty results.
• Filter Limitations: Only supported filters (keywords, time range, participants, organizers, meeting rooms) are available. Unsupported filters must be communicated to the user.
• Pagination: Search results may span multiple pages. Always retrieve all pages to avoid missing records.
• Artifact Directory Structure: All files for a meeting (cover, notes, transcript) must reside in a single ./artifact-<title>/ directory for consistency.
• Whiteboard Cover: Not all meetings will have a whiteboard cover image. Handle this case gracefully.
• Do Not Read Document Content by Default: Only fetch and display document content (summary, to-do, sections) if the user explicitly requests it. Normally, provide just the links.

By following these guidelines, you can maximize the utility of the vc (v1) skill for Feishu meeting management on the Happycapy Skills platform.