refinement

Pass

Audited by Gen Agent Trust Hub on Nov 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted materials from the ai/inputs/ directory, which may include text, documents, and external links. These materials are used to generate feature specifications, but the process lacks safety boundaries or sanitization, creating a surface for malicious instructions to influence the agent's behavior.
  • Ingestion points: Reads from ai/inputs/ and external URLs referenced within those inputs (SKILL.md).
  • Boundary markers: Absent; the instructions do not provide delimiters or warnings to ignore embedded instructions within the input materials.
  • Capability inventory: The skill performs filesystem write operations to create or update files in the ai/features/ directory (SKILL.md).
  • Sanitization: Absent; there is no mention of validation or filtering for the content gathered from external sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 22, 2026, 11:06 AM