skills/byronfinn/dev-skills/tdd/Gen Agent Trust Hub

tdd

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources including issue bodies and PRD files. This creates a surface for indirect prompt injection where malicious instructions could be embedded in requirements.
  • Ingestion points: Requirements are loaded from Issue references and PRD files (SKILL.md Step 1, REFERENCE.md Chapter 1).
  • Boundary markers: The skill uses human review gates (Scenario Review) to confirm criteria, which serves as a manual boundary.
  • Capability inventory: The skill is capable of reading local documentation and writing/executing code.
  • Sanitization: No specific content sanitization or instruction-ignoring delimiters are mentioned for the ingested text.
  • [DYNAMIC_EXECUTION]: The skill generates and executes code (tests and implementation) as its primary workflow.
  • Evidence: The cycle described in SKILL.md and REFERENCE.md (Chapters 3 and 4) involves writing test files and implementation files and running them to verify behavior.
  • Control: This risk is managed through mandatory human checkpoints (Test Code Review Gate) where the user reviews the generated code before it is finalized.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 06:21 AM
Security Audit — agent-trust-hub — tdd