yt-plan
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill reads from a specific configuration file path
~/.claude/youtube/voice.mdto personalize its recommendations. This is a targeted access to a local application-specific configuration file and does not involve accessing sensitive system credentials or private keys. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data by transcribing user-provided videos to infer style and tone. While this represents a data ingestion surface, the risk is minimal as the skill's output is limited to text-based planning tables and it does not possess capabilities for file modification, network exfiltration, or code execution.
- Ingestion points: Transcriptions of user-provided videos and the voice profile file.
- Boundary markers: None explicitly defined in instructions.
- Capability inventory: No subprocess calls, network operations, or file-write capabilities are requested or used.
- Sanitization: Standard LLM processing is applied to transcriptions without specific regex or filtering logic.
Audit Metadata