Transcribee 🐝

Transcribe YouTube videos and local audio/video files with speaker diarization. Use when user asks to transcribe a YouTube URL, podcast, video, or audio file. Outputs clean speaker-labeled transcripts ready for LLM analysis.

Before installing or enabling this skill: 1) Expect it to require two API keys (ELEVEN_LABS_API_KEY and ANTHROPIC_API_KEY) and system binaries (yt-dlp, ffmpeg) even though the registry listing omitted them — verify and supply keys only if you trust those services. 2) Be aware audio and transcripts are uploaded to external services; if privacy-sensitive audio will be transcribed, consider running a local-only alternative. 3) The skill will read your ~/Documents/transcripts/ library and write new transcript folders there — review or sandbox it if you don't want that folder modified. 4) Verify the skill's source (there is no homepage listed) — prefer installing from a trusted repo and inspect the .env.example and index.ts for any extra endpoints or hardcoded secrets. 5) If you allow autonomous agent invocation, consider restricting its access or running the skill manually the first few times to confirm behavior. If you want to go ahead, run it in an isolated environment, provide least-privilege API keys, and review the code for any hidden network calls not documented in README/CLAUDE.md.

Type: OpenClaw Skill Name: transcribee Version: 1.2.1 The skill is classified as benign. It transparently uses `yt-dlp` and `ffmpeg` for media processing and `ElevenLabs` and `Anthropic` APIs for transcription and categorization, which aligns with its stated purpose. External command execution is handled using `execFileAsync`, which is a safer method than `exec` as it prevents shell injection. Output files are saved to a user-owned directory (`~/Documents/transcripts`). There is no evidence of data exfiltration, malicious execution, persistence mechanisms, or prompt injection attempts against the OpenClaw agent itself; the prompt engineering observed is for the internal Anthropic LLM used for categorization.