report-distribution-agent
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a consolidation agent and territory queries, representing a standard architectural surface for indirect prompt injection common to data-processing agents. 1. Ingestion points: Data generated by the Data Consolidation Agent and territory database queries as defined in the Workflow Process. 2. Boundary markers: The skill instructions do not explicitly specify delimiters to isolate ingested data from agent instructions. 3. Capability inventory: The skill is equipped with SMTP transport capabilities and local execution via a runner script. 4. Sanitization: No explicit sanitization or validation of the consolidation agent's output is mentioned in the workflow steps.
- [COMMAND_EXECUTION]: The skill uses a standard bash helper script to invoke a local Python-based runner. Evidence: scripts/run.sh executes a runner script (run-skill.py) using a relative path within the repository environment.
Audit Metadata