result-integrator
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill processes untrusted input from external analysis sources. However, because its capabilities are strictly restricted to the 'Read' tool and it lacks execution or network functionality, the risk is minimal.
- Ingestion points: The skill integrates 'Plot analysis results' from multiple sources.
- Boundary markers: No delimiters are specified to isolate external data.
- Capability inventory: Only the 'Read' tool is permitted; no file-write, network, or command execution tools are present.
- Sanitization: No input validation is mentioned.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or access to sensitive local file paths were identified.
- [Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill does not reference external packages or remote scripts.
Audit Metadata