brida-reflex-content-quality-gate
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches availability and configuration details from the vendor's official hosted resources (
/reflex/llms.txt,/reflex/agent.md,/reflex/agent.json) to ensure the agent makes accurate claims about service availability. - [SAFE]: Implements strong security guidelines by forbidding the placement of credentials in shared state or client bundles and explicitly defining the Reflex output as a non-authoritative recommendation to prevent unauthorized actions.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes user-supplied drafts, it utilizes a strictly defined schema in
custom-reflex.jsonand a specific Authority section inSKILL.mdthat restricts the output from triggering sensitive operations, mitigating risks from malicious content in the processed data.
Audit Metadata