flow-brainstorm
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read external source code or supplied artifacts to ensure architectural compatibility during brainstorming. This represents a potential data ingestion point for indirect instructions from those artifacts.
- Ingestion points: Referenced in the process section to read "relevant source or supplied artifacts" from the user's project.
- Boundary markers: Not explicitly defined in the skill instructions.
- Capability inventory: The skill contains no executable code, shell commands, file-write operations, or network requests.
- Sanitization: No explicit sanitization or filtering of external content is described.
- [SAFE]: The skill consists entirely of natural language instructions for collaborative idea exploration. It contains no hardcoded credentials, obfuscated content, persistence mechanisms, or unauthorized commands.
Audit Metadata