pipeline-update
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external CRM systems (deal names, company info, notes) which could contain malicious instructions.
- Ingestion points: CRM deal records retrieved during the 'Look up existing deal record' step.
- Boundary markers: None present to distinguish between data and instructions.
- Capability inventory: Executes a local Python script (approval-manager.py), writes to log files (pipeline-update-log.json), and performs CRM updates.
- Sanitization: No explicit sanitization or validation of the retrieved CRM content before processing is mentioned.
- [COMMAND_EXECUTION]: The skill instructions direct the agent to execute a specific local Python script (approval-manager.py) to manage approval logging. While this is part of the skill's operational logic, it represents a command execution capability.
Audit Metadata