design-taxonomy

Pass

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill demonstrates an attack surface for indirect prompt injection because it ingests untrusted data from a user and passes it to downstream components without explicit sanitization or boundary markers.
  • Ingestion points: User responses gathered via the AskUserQuestion tool in Step 2.
  • Boundary markers: Absent. User responses are interpolated directly into the invocation of the taxonomy-architecture skill.
  • Capability inventory: The skill has access to Read, Glob, Grep, Task, Skill, and AskUserQuestion tools.
  • Sanitization: Absent. There is no evidence of escaping or validating the input before passing it to the next skill.
  • [Unverifiable Dependencies & Remote Code Execution] (SAFE): No remote scripts are downloaded or executed. The skill only uses internal tools and generates static code templates.
  • [Data Exposure & Exfiltration] (SAFE): The skill does not access sensitive file paths (e.g., credentials, SSH keys) or perform network operations to external domains.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 18, 2026, 05:46 AM