trigger-refactor-pipeline
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (LOW): The script analyze_trigger.py ingests Apex code from a Salesforce org. Evidence: 1. Ingestion point: scripts/analyze_trigger.py via subprocess call to 'sf apex get'. 2. Boundary markers: None present to delimit untrusted code from analysis output. 3. Capability inventory: Bash tool, Read/Write tools, and Salesforce deployment commands. 4. Sanitization: None; the script uses regex directly on external source code. Malicious instructions embedded in Apex comments could influence agent behavior.
- SAFE (SAFE): No hardcoded credentials, obfuscation, persistence mechanisms, or unauthorized network exfiltration were detected. The skill performs standard software engineering tasks.
Audit Metadata