afa-scale
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest raw user data through the
user_requestfield and pass it to downstream workers (afa-ops,afa-expand). This creates an attack surface where instructions embedded in user data could potentially influence agent behavior across the multi-step coordination workflows. \n - Ingestion points: The
user_requestfield defined in the input contract within SKILL.md. \n - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are defined for the
user_requestinput. \n - Capability inventory: The skill has the ability to route tasks to multiple sub-workers and write profile/asset data to the
./brand-brain/directory. \n - Sanitization: There is no evidence of input validation or sanitization protocols for the external content before it is processed or handed off.
Audit Metadata