updating-knowledge
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch and process content from external URLs using the
web_fetchandweb_searchtools. This creates an attack surface where an attacker-controlled website could include instructions designed to override the agent's behavior or manipulate the research summary. - Ingestion points: External content fetched via
web_searchandweb_fetch(SKILL.md). - Boundary markers: The skill instructions emphasize internal synthesis in "thinking blocks" and cross-validation across at least 3 sources (SKILL.md).
- Capability inventory: The skill primarily performs read operations and synthesizes information. It does not appear to have file-write or command-execution capabilities.
- Sanitization: The skill relies on "Synthesis" and "Source Gathering" protocols to filter and cross-validate information, but does not explicitly detail technical sanitization of fetched text.
Audit Metadata