ap2-intent-mandate
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to fetch and process documentation from external websites (
ap2-protocol.org) and search results to guide its execution. - Ingestion points: Content is retrieved from external URLs using the
WebFetchtool. - Capability inventory: The agent is configured with high-capability tools, including
Bash,Write, andEdit. - Boundary markers: The instructions lack boundary markers or warnings to disregard potential instructions embedded within the fetched documentation.
- Sanitization: No sanitization or validation of the external content is requested before the agent uses it to inform its actions.
- [EXTERNAL_DOWNLOADS]: The skill fetches protocol documentation and terminology from
ap2-protocol.org. These downloads are documented as necessary for the skill's primary purpose of implementing the AP2 Intent Mandate protocol.
Audit Metadata