requirement-refine

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of raw requirements (text, tickets, emails) and multimedia assets (images, diagrams). It also utilizes a subagent for validation, creating a surface where malicious instructions embedded in the input could influence the agent's behavior or the validation results.
  • Ingestion points: Raw requirements and user-provided files (images, PDFs, captures) are stored in the 'docs/specs/requirements/SRS-XXX-[nombre-corto]/references/' directory and read during the specification process.
  • Boundary markers: The instructions describe a structured workflow but do not explicitly mandate the use of delimiters or specific 'ignore' instructions when the ingested content is processed or passed to the validation subagent in Ola 3.
  • Capability inventory: The skill has the capability to write files to the local filesystem (spec documents and wireframes) and can invoke other agent skills such as '/work-research', '/work-define', and '/work-plan'.
  • Sanitization: There is no evidence of explicit sanitization, filtering, or escaping of the ingested content before it is interpolated into the specification document or provided to the validation subagent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 08:06 PM
Security Audit — agent-trust-hub — requirement-refine