synthesize-snapshots
Pass
Audited by Gen Agent Trust Hub on Mar 17, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No evidence of instructions designed to bypass safety filters or override agent behavior. The guidelines are strictly task-oriented.- [DATA_EXPOSURE_AND_EXFILTRATION]: No access to sensitive system files or credentials detected. The skill operates on local project directories for research data.- [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: No remote code execution or external package installations are requested.- [INDIRECT_PROMPT_INJECTION]: The skill processes interview snapshots which are external data. While this presents an ingestion surface, the risk is inherent to the primary purpose of data synthesis and is mitigated by the structured template requirements.
Audit Metadata