swift-tdd

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): The skill uses authoritative language to enforce TDD principles (e.g., 'The Iron Law'), which is appropriate for its educational purpose and does not attempt to bypass AI safety guards.
  • [Data Exposure & Exfiltration] (SAFE): No evidence of hardcoded secrets, sensitive file path access, or unauthorized network requests. All code examples are standard Swift unit testing logic.
  • [Remote Code Execution] (SAFE): No patterns of downloading or executing remote scripts were found.
  • [Obfuscation] (SAFE): The content is clear and uses standard Markdown and Swift syntax with no hidden or encoded characters.
  • [Indirect Prompt Injection] (SAFE): Analysis of vulnerability surface: (1) Ingestion points: User-provided Swift code or project files. (2) Boundary markers: None explicitly defined. (3) Capability inventory: Code generation and refactoring; no shell tools defined within the skill. (4) Sanitization: Relies on standard LLM output safety.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 03:58 PM