tdd-workflows-tdd-green

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill provides a structured process for an agent to implement code based on test failure data.
  • Ingestion points: Failing test results are passed as input via the $ARGUMENTS placeholder in the implementation prompt within resources/implementation-playbook.md.
  • Boundary markers: Data is interpolated into the sub-agent's prompt as part of the implementation instructions.
  • Capability inventory: The skill uses a specialized sub-agent (test-automator) to generate code modifications.
  • Sanitization: The process relies on the internal reasoning of the agent to handle the provided test data.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 02:47 AM