TG Voice Whisper Transcriber
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill fits a local Telegram voice transcriber, but it asks to create persistent automation that can automatically send transcriptions and delete voice files.
Install only if you intentionally want always-on Telegram voice transcription. Before enabling auto-setup, add a clear stop/removal process, restrict watched paths and target chats, validate file paths, and consider requiring confirmation before sending transcriptions or deleting local files. Use an isolated Python environment instead of --break-system-packages where possible.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A mistaken or unexpected file path could lead to unwanted message sends or local file deletion.
This directs the agent to run a raw shell chain that transcribes a file, reads wildcard output, sends a chat message, and removes files, but it does not define safe path handling, confirmation, or rollback behavior.
exec 'whisper PATH --model tiny --language ru --output_format txt --output_dir /tmp/whisper; text=$(cat /tmp/whisper/*.txt); message action=send ...; rm PATH /tmp/whisper/*'
Require explicit user approval before sending/deleting, quote and validate file paths, use a dedicated temporary directory, and handle failures before running rm.
The automation may continue scanning, replying, and deleting files even when the user no longer expects it.
The skill asks for persistent background automation that keeps checking for files and can keep running after setup, with no clear stop or cleanup instruction.
**Auto-setup (non-block):** Spawn sub-agent or cron every 5s ... sessions_spawn ... label="voice-auto" cleanup="keep"
Only enable this with explicit consent, document how to disable/remove the cron or spawned session, and consider making transcription user-invoked rather than always-on.
Installation could change the local Python environment or pull whatever package version is current at install time.
The skill depends on external package installation, including an unpinned pip package and a flag that can modify system-managed Python packages. This is related to the stated local Whisper purpose but should be reviewed by the user.
install:
- kind: apt
packages: [ffmpeg]
- kind: pip
packages: [openai-whisper]
flags: [--break-system-packages]Prefer a virtual environment or container, pin dependency versions, and make installation requirements visible in the registry metadata.