p-video
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: Accesses and manages the PIXAZO_API_KEY. The skill checks environment variables, reads from a dedicated local file (~/.pixazo/api-key), and provides instructions for the user to provide the key if not found.
- [COMMAND_EXECUTION]: Instructs the agent to use chmod 600 when saving the API key to the filesystem. This is a standard security practice to ensure only the current user can access the sensitive file.
- [EXTERNAL_DOWNLOADS]: Fetches additional model specifications and polling instructions from the vendor's official domain (https://www.pixazo.ai/models/p-video.md) to ensure correct API interaction.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill fetches model reference data and polling logic from https://www.pixazo.ai/models/p-video.md.
- Boundary markers: None explicitly defined for the remote content ingestion.
- Capability inventory: Includes file system read/write access (~/.pixazo/api-key), network requests (curl, Python requests, Node fetch), and shell command execution (chmod).
- Sanitization: None described for the external content fetched from the vendor website.
Audit Metadata