← Back to Skills Marketplace
110
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install meeting-secretary
Description
专业的会议秘书技能，用于分析会议转录文本并生成结构化会议纪要。当用户提供会议转录文本（或会议笔记文本），并要求分析会议内容、生成会议纪要、提取关键信息时使用此技能。
Usage Guidance
This skill appears coherent and implements what it claims: analyze meeting transcripts and produce structured minutes. Things to consider before installing/use:
- The included script (scripts/split_long_transcript.py) is a simple, readable helper that splits long transcripts and writes segment files to disk—expect files to be created in the specified output directory. Review and run it in a safe workspace (not as root).
- Documentation references additional scripts (transcribe_audio.py, process_video.py) and integrations (Notion/飞书) that are NOT included. If you need audio/video processing or automatic uploads, you'll need to obtain or implement those scripts and supply any API keys yourself.
- The README suggests using packages like whisper or moviepy for optional features; those are not installed automatically. Installing those packages and running transcription will have their own requirements and privacy implications.
- The skill processes potentially sensitive meeting content. Do not feed confidential transcripts unless you are comfortable with how outputs and any local files are stored and with any downstream services you integrate.
Recommendation: you can install and use this skill for text-only transcript analysis after reviewing the script. If you plan to use audio/video processing or service integrations, verify those additional tools and required credentials separately and inspect any third-party code before running.
Capability Analysis
Type: OpenClaw Skill
Name: meeting-secretary
Version: 1.0.0
The 'meeting-secretary' skill bundle is a legitimate tool designed for analyzing meeting transcripts, identifying participant sentiments, and generating structured minutes. The provided Python script, `scripts/split_long_transcript.py`, is a utility for segmenting large text files based on timestamps or speakers using standard libraries and safe file handling practices. While the documentation in `references/best_practices.md` refers to additional scripts (transcribe_audio.py, process_video.py) not included in the bundle, the existing code and the AI instructions in `SKILL.md` focus entirely on the stated analytical purpose without any evidence of data exfiltration, malicious execution, or unauthorized system access.
Capability Assessment
Purpose & Capability
Name/description describe transcript analysis and structured minutes; included SKILL.md and README focus on deep textual analysis and dynamic summary generation. The repository includes a legitimate helper script (scripts/split_long_transcript.py) that aligns with the stated need to split long transcripts. There is no request for unrelated credentials or binaries.
Instruction Scope
Runtime instructions tell the agent to read user-provided transcript text, optionally run scripts/split_long_transcript.py for long texts, analyze segments, and produce structured summaries—this stays within the expected scope. A documentation note: SKILL.md and references mention other scripts (transcribe_audio.py, process_video.py) and integrations (Notion/Feishu) that are not present as code or declared environment requirements; these are optional suggested workflows but not part of the included runtime instructions.
Install Mechanism
No install spec is provided (instruction-only with a small helper script), so nothing will be auto-downloaded or executed during install. The included Python script is plain and readable; it writes segment files when run (expected behavior).
Credentials
The skill declares no required environment variables or credentials. README mentions third-party tools and optional integrations (Whisper, Notion/飞书/Obsidian) that would require additional packages or API keys if the user chooses to use them—those credentials are not requested by the skill itself. Users should be aware that enabling suggested integrations would require granting external service credentials separately.
Persistence & Privilege
Skill does not request permanent/always-on presence (always:false), does not modify other skills or system-wide settings, and does not declare elevated privileges. It reads user-supplied transcripts and can write segment files to an output directory when run (normal and expected).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install meeting-secretary - After installation, invoke the skill by name or use
/meeting-secretary - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
meeting-secretary 1.0.0 – 专业会议纪要智能分析技能首次发布
- 支持自动识别会议类型（如决策、讨论、项目、头脑风暴等）并自适应纪要结构
- 深度分析会议转录文本，精准提取观点、立场、分歧、共识及待办事项
- 提供细粒度的语气、语态、情感、潜台词及高价值信息挖掘
- 高度结构化的纪要输出，覆盖决策过程、分歧焦点、行动计划等核心
- 集成智能文本分段工具，兼容长文本场景
- Markdown格式输出，便于通用存档和后续追溯
Metadata
Frequently Asked Questions
What is meeting-secretary?
专业的会议秘书技能，用于分析会议转录文本并生成结构化会议纪要。当用户提供会议转录文本（或会议笔记文本），并要求分析会议内容、生成会议纪要、提取关键信息时使用此技能。 It is an AI Agent Skill for Claude Code / OpenClaw, with 110 downloads so far.
How do I install meeting-secretary?
Run "/install meeting-secretary" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is meeting-secretary free?
Yes, meeting-secretary is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does meeting-secretary support?
meeting-secretary is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created meeting-secretary?
It is built and maintained by Carter Wu (@wuzhiguocarter); the current version is v1.0.0.
More Skills