sast-businesslogic
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted codebase content and architectural summaries as its primary data source. This creates a surface for indirect prompt injection, where an attacker could place malicious instructions in code comments or documentation to influence the agent's findings or behavior.
- Ingestion points: The skill ingests codebase files and the
sast/architecture.mdfile to identify business features and generate attack scenarios. - Boundary markers: The skill defines clear goals for its subagents but does not include explicit delimiters or "ignore embedded instructions" warnings for the external data it processes.
- Capability inventory: The skill is capable of reading from and writing to the file system (specifically the
sast/directory) and orchestrating subagent tasks. - Sanitization: No sanitization or filtering is applied to the analyzed code to prevent the interpolation of malicious instructions into the subagent prompts.
Audit Metadata