ClawHub

Speech to Text Transcription

v1.0.0

Transcribe audio and video files to text with speaker detection, timestamps, and format conversion.

⭐ 0· 736·3 current·3 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description match the behavior: audio/video processing with ffmpeg, local Whisper or cloud providers (OpenAI, AssemblyAI, Deepgram). No unrelated binaries, env vars, or endpoints are requested.
ℹ
Instruction Scope
Instructions cover verifying local files, downloading URLs to a temp folder, splitting/processing with ffmpeg, calling cloud transcription APIs only when chosen, and saving transcripts. A potential privacy note: the skill explicitly stores transcripts and a 'memory' file (provider preferences, usage patterns) under ~/speech-to-text-transcription/, and the guidance to 'learn from what they transcribe' could lead to persistent storage of sensitive metadata unless the user is explicit; otherwise the runtime instructions stay within the stated purpose.
✓
Install Mechanism
This is instruction-only: there is no install script or arbitrary download. The SKILL.md suggests installing local Python Whisper via pip if desired, which is proportional and optional. No remote archives or opaque installs are invoked by the skill itself.
✓
Credentials
No required environment variables. Optional API keys (OPENAI_API_KEY, ASSEMBLYAI_API_KEY, DEEPGRAM_API_KEY) are appropriate and proportional to enabling cloud providers; they are optional and documented.
ℹ
Persistence & Privilege
The skill writes to its own directory in the user's home and persists transcripts and a memory.md file. It does not request system-wide privileges, nor is always:true set. Users should be aware of on-disk persistence of possibly sensitive content.
Assessment
This skill appears coherent for transcription, but consider these before installing: (1) It will create ~/speech-to-text-transcription/, store transcripts there, and keep a memory.md with preferences—delete or encrypt that directory if you are transcribing sensitive audio. (2) Cloud transcription (OpenAI, AssemblyAI, Deepgram) will upload audio to third parties only if you choose those providers and set API keys; for sensitive material prefer local Whisper. (3) The skill may download URLs and write temp files during processing—ensure you trust source URLs. (4) The skill is instruction-only (no install script), but it suggests optionally running pip install openai-whisper to enable local transcription. (5) Confirm the agent prompts you before uploading any audio and review/clean the stored transcripts and memory if privacy is a concern.

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

latestvk97612eaeydknpkr2zhv2aah1581nmcq

License

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

Runtime requirements

🎤 Clawdis
OSLinux · macOS · Windows
Binsffmpeg

Comments