vc-outreach
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied startup data to populate email templates, creating a surface for indirect prompt injection.\n
- Ingestion points: User-provided startup metrics and descriptions in SKILL.md.\n
- Boundary markers: None present.\n
- Capability inventory: Includes local file reading via scripts/check-output.py; no network or write access.\n
- Sanitization: No content filtering is defined.\n- [COMMAND_EXECUTION]: The skill includes scripts/check-output.py, which performs benign keyword checking on agent-generated files.
Audit Metadata