dev-workflow

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes npm test -- --coverage in Phase 6 to run test suites and verify code coverage. This is a standard operation within a development workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of PRDs and GitHub issues. This creates a surface where instructions embedded in project requirements could influence agent behavior.
  • Ingestion points: PRD issues and task descriptions retrieved from GitHub (SKILL.md).
  • Boundary markers: None explicitly defined in the prompt instructions to separate data from instructions.
  • Capability inventory: File system modification (Phase 5 development), shell execution (npm test), and GitHub API interactions for PR and issue management (Phase 8).
  • Sanitization: Not explicitly implemented in the skill logic; the workflow relies on a human-led Code Review (Phase 7) as a security gate.
  • [DYNAMIC_EXECUTION]: Phase 6b describes an automated Playwright flow involving a 'Generator' that writes tests and a 'Healer' that automatically patches them. This involves runtime script generation and execution of code based on the current implementation state.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 02:53 PM
Security Audit — agent-trust-hub — dev-workflow