yt-plan
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external video content and a local configuration file, which could contain malicious instructions designed to influence agent behavior.
- Ingestion points: Reads voice profile data from
~/.claude/youtube/voice.mdand transcribes user-provided YouTube videos. - Boundary markers: Absent; there are no instructions for the agent to treat ingested content as data only or to ignore embedded instructions.
- Capability inventory: The agent is instructed to read from and write to the local file system using configuration files.
- Sanitization: Absent; no validation or filtering of ingested video transcripts or profile content is specified.
- Ingestion points: Reads voice profile data from
- [DATA_EXFILTRATION]: The skill instructions involve reading from and writing to a specific directory in the user's home folder.
- Evidence: The skill explicitly directs the agent to read from and write to
~/.claude/youtube/voice.md. Although this is a configuration path for the agent, accessing files in the home directory is a potential data exposure surface.
- Evidence: The skill explicitly directs the agent to read from and write to
Audit Metadata