camp-profile
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill drafts information from an external website provided by the user, which presents a surface for indirect prompt injection.
- Ingestion points: External website URL provided by the director.
- Boundary markers: No explicit delimiters are used for the website content.
- Capability inventory: Creating and editing .agents/camp-profile.md on the local filesystem.
- Sanitization: The agent is instructed to walk the director through the draft for confirmation and correction.
- [SAFE]: The skill performs expected file system operations to store durable facts in a shared configuration file, following the platform's conventions for agent context.
Audit Metadata