cv-generator
Pass
Audited by Gen Agent Trust Hub on Aug 5, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to ingest and process content from external sources, which presents a surface for indirect prompt injection where malicious instructions embedded in external data could attempt to subvert the agent's logic.
- Ingestion points: Data enters the context through LinkedIn profile URLs, personal portfolio/website URLs, and uploaded resume files (PDF, DOCX, TXT) as specified in Sections 2 and 4.
- Boundary markers: The instructions include an 'Anti-hallucination enforcement gate' (Section 10) and 'Tense enforcement' (Section 7d) to validate output against source data. While robust for quality control, these do not explicitly instruct the model to ignore adversarial commands embedded in the source text.
- Capability inventory: The skill requires capabilities for network fetching (URLs), document parsing, and potential OCR processing as noted in Section 2.
- Sanitization: The skill implements a 'machine-checkable blocklist' for banned phrases (Section 7c) and a truthfulness score (Section 11e), providing a layer of validation for the generated content.
Audit Metadata