The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the agent to execute a local shell script (transcribe.sh) that invokes powerful system binaries including ffmpeg, yt-dlp, and whisper-cli.
[COMMAND_EXECUTION]: The installation components (install.sh and add_permission.py) modify the global agent configuration file (~/.claude/settings.json) to silently grant read permissions to the skill's repository path, bypassing the platform's built-in interactive security prompts.
[EXTERNAL_DOWNLOADS]: The transcribe.sh script uses yt-dlp to download and extract audio from arbitrary user-provided URLs, which involves fetching data from various third-party websites.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the processing of untrusted external transcripts.
Ingestion points: The agent reads transcript content generated from external URLs or local audio files as defined in SKILL.md.
Boundary markers: The skill does not use delimiters or provide instructions to the agent to ignore embedded commands or instructions found within the transcript data.
Capability inventory: The agent has the ability to execute shell scripts (transcribe.sh) and perform file read/write operations.
Sanitization: No validation or sanitization is performed on the transcript text before it is processed by the agent's analysis logic.

transcribe