bmad-agent-dev

Story implementation with strict test-driven development executes approved tasks in exact sequence, requiring comprehensive unit test coverage and a fully passing test suite before marking any item complete. It updates the story file with an auditable record of decisions and changed files for traceability.

Agentic governance in software development workflows ensures an agent executes tasks exactly as written in a story file without reordering or skipping. It gates each task completion on real test execution outcomes and records an auditable Dev Agent Record back into the story for traceability.

Yes, test-first completion gating requires the full test suite to pass before marking items complete. The implementation process demands comprehensive unit tests run after each subtask, ensuring verifiable passing test results gate every step of story execution.

Maintaining an auditable file list during code implementation requires the agent to record decisions, added tests, and updated file lists directly back into the story file. This creates traceable development records tied to acceptance criteria.

Story execution can still proceed if project context is missing, but it loads bmad initialization config and project context when available to inform the process. The agent strictly gates task completion on real test execution outcomes regardless of context.

You should not use automated story execution when tasks and acceptance criteria are not provided via an approved story file. The process relies on strict sequence execution and real test outcomes, making it unsuitable for unstructured development without predefined test gates.