Draft Camp Emails
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [DATA_EXPOSURE]: The skill reads from 'family-profile.md', which contains sensitive personal information such as children's names, birth dates, and medical or dietary restrictions. This data is used solely to personalize email templates locally and is not exfiltrated to external servers.
- [INDIRECT_PROMPT_INJECTION]: The skill processes camp provider research files that may contain untrusted data from web sources, creating a potential vector for indirect instructions.
- Ingestion points: Camp provider details are read from files located at '<research_dir>/providers/[provider].md'.
- Boundary markers: The instructions do not provide delimiters or specific warnings to ignore instructions that might be embedded within the research data.
- Capability inventory: The skill has the capability to read project-specific configuration and research files and write drafted markdown files to the local filesystem.
- Sanitization: Content from external provider files is interpolated into email templates without explicit validation or escaping of the content.
Audit Metadata