ClawHub

调用senseaudio asr的课堂转译助手，将英文课堂录音、讲座视频、组会音频等内容自动转写为文本，生成中文总结

v1.0.1

接收课堂录音、讲座音频或视频文件（视频会先抽取音轨），调用 SenseAudio HTTP ASR API 进行英文转录，可选直出中文翻译；随后整理为结构化 Markdown 学习笔记，包含摘要、关键概念、术语表、时间轴与复习问题，生成到桌面，并支持导出到 Notion 或保存到 Obsidian vault。

⭐ 0· 125·0 current·0 all-time
by@klilyz
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
medium confidence
✓
Purpose & Capability
Name/description match what the files do: script accepts audio/video or a page URL, extracts audio (ffmpeg for video), calls the SenseAudio HTTP ASR endpoint, optionally calls OpenAI-compatible summary provider and/or Notion API, and writes Markdown/Obsidian files. Required binary (python3) and declared env SENSEAUDIO_API_KEY are appropriate for the stated functionality.
ℹ
Instruction Scope
SKILL.md and script instruct the agent to fetch page HTML (when --page-url is used), download media, extract audio, upload the audio to SenseAudio, and optionally push content to Notion or write into an Obsidian vault. These actions are within the stated scope but have privacy implications: user audio (and any media downloaded from a page) will be transmitted to external services (SenseAudio and optionally Notion/OpenAI). The instructions do not appear to read unrelated local config or credentials.
✓
Install Mechanism
No install script that pulls arbitrary binaries; dependencies are minimal (requests). This is an instruction-only skill with a Python script; risk from install mechanism is low. Users must manually install dependencies and ffmpeg for video handling per docs.
✓
Credentials
Primary required env var is SENSEAUDIO_API_KEY, which is justified. OPENAI_API_KEY, OPENAI_BASE_URL, and NOTION_TOKEN are optional and only needed for explicit optional functionality. No unrelated credentials are requested.
✓
Persistence & Privilege
always is false and the skill does not request permanent platform presence or attempt to modify other skills. It writes output files to user-specified directories and optionally to a supplied Obsidian vault path, which is expected behavior.
Assessment
This skill behaves as advertised but you should be aware of privacy and data-flow consequences before installing: any audio you provide (or media downloaded from a web page via --page-url) will be uploaded to SenseAudio for transcription, and summaries may be sent to an OpenAI-compatible endpoint if you enable that provider; exporting to Notion requires giving a Notion token which grants the skill access to create pages. If your recordings contain sensitive or confidential information, do not run them through this skill or review SenseAudio/Notion/OpenAI privacy and retention policies first. Also ensure ffmpeg is installed locally for video inputs and that the output directory / Obsidian vault paths are correct and writable. If you want higher assurance, review the full script locally (scripts/classroom_note_translator.py) before running — the code is small and readable and the scanned package shows no pre-scan injection signals.

Like a lobster shell, security has layers — review code before you run it.

latestvk97337zm2gjvbj313q3kmt5t65833dc4

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Runtime requirements

🎓 Clawdis
Binspython3
EnvSENSEAUDIO_API_KEY
Primary envSENSEAUDIO_API_KEY

Comments