brida-reflex-catalog-listing-signals
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch updated implementation guidance and API availability information from the vendor's official resources, including
/reflex/llms.txt,/reflex/agent.md, and/reflex/agent.json. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and classify external catalog data (titles and descriptions). The risk of data-driven prompt injection is mitigated by the skill's explicit authority boundary, which restricts findings to recommendations and prohibits them from granting permissions for protected actions.
- [SAFE]: The instructions include security best practices, such as requiring revalidation of authorization before any side effects and explicitly forbidding the storage of reusable credentials in the classification state.
Audit Metadata