derive

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted data from specification files to generate executable code artifacts, creating a surface for indirect prompt injection.
  • Ingestion points: Uses the Read tool to ingest content from SPEC-XXX.md files as described in the workflow.
  • Boundary markers: Relies on instructional Anti-Hallucination Rules to bound the output, but lacks technical delimiters or sanitization for input data.
  • Capability inventory: Possesses the capability to search (Grep/Glob) and create files (Write) such as .feature and .test.ts files.
  • Sanitization: No automated sanitization or validation of the input specification content is performed before the AI generates code from it, relying entirely on manual human review.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 03:41 PM