research
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is designed for research and documentation purposes, with no executable code or malicious patterns detected.
- [INDIRECT_PROMPT_INJECTION]: The skill has an inherent attack surface for indirect prompt injection as it is designed to ingest and process untrusted external data.
- Ingestion points: Official documentation, source code, and APIs (referenced in SKILL.md).
- Boundary markers: No delimiters or ignore instructions are specified for external content.
- Capability inventory: The agent has the capability to write research findings into Markdown files in the local repository (SKILL.md).
- Sanitization: No explicit sanitization of the external content is mentioned.
Audit Metadata