skill-profile-builder
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has a potential attack surface for indirect prompt injection because it fetches and processes content from external social media URLs using the
web_fetchtool. - Ingestion points: Social media links provided by the user are fetched via
web_fetchinreferences/social-link-analysis.md. - Boundary markers: The instructions lack specific delimiters for untrusted data, but they require the agent to cross-check all inferred conclusions with the user.
- Capability inventory: The skill performs file-write operations to the
profiles/directory. - Sanitization: No explicit sanitization of external content is mentioned, though the skill relies on the agent's internal reasoning and user confirmation steps.
- [DATA_EXPOSURE]: The skill writes user-provided data and analyzed content to local files within the
profiles/directory. Access is limited to the defined folder structure, and the skill explicitly instructs against crawling private data or bypassing login walls.
Audit Metadata