to-conformance

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes normative requirements from external or linked specification documents which may contain adversarial content.
  • Ingestion points: Reads specifications from spec-path, external linked documents, and the current conversation context.
  • Boundary markers: No specific delimiters or "ignore embedded instructions" guards are implemented for the ingested specification content.
  • Capability inventory: The skill has the capability to write to the local file system (defaulting to docs/CONFORMANCE.md).
  • Sanitization: Input from specifications is processed directly to generate implementation-independent tests without explicit sanitization or validation of the source text's intent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 06:17 PM
Security Audit — agent-trust-hub — to-conformance