disciplined-quality-evaluation
Warn
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (MEDIUM): The skill is vulnerable to Indirect Prompt Injection (Category 8). \n- Ingestion points: Processes untrusted 'Research Documents' and 'Implementation Plans' in 'Step 1: Document Intake'. \n- Boundary markers: No explicit delimiters (e.g., XML tags) or instructions to disregard embedded commands are present in the instruction set. \n- Capability inventory: The skill does not contain executable scripts or network access, but it issues 'GO/NO-GO' decisions intended to 'block phase transitions', creating a significant side-effect surface in automated development workflows. \n- Sanitization: No sanitization, escaping, or schema validation is applied to the ingested content before evaluation.
Audit Metadata