using-my-poor-ai

To enforce TDD and structured engineering discipline in AI coding sessions, this Skill classifies development requests into specific paths and mandates verification protocols, ensuring agents perform root-cause analysis instead of superficial fixes.

The best way to orchestrate multiple AI subagents for complex software development is by triggering a brainstorming phase that generates a design document and coordinates agents to execute the plan in parallel, ensuring architectural integrity.

To prevent AI coding agents from applying superficial fixes instead of root-cause debugging, you need an orchestration pipeline that mandates systematic debugging workflows and classifies requests based on complexity before execution.

To verify that an AI coding agent has met defined success criteria before task completion, enforce verification guardrails that mandate the creation of GOAL.md files and explicit verification steps prior to completion.

Yes, you can use automated pipeline orchestration for both simple and complex feature development because the Skill automatically classifies requests into Debug, Simple, or Full development paths based on the complexity of the task.

You need architectural planning and design documents for AI-assisted coding when starting a complex feature, which triggers a brainstorming phase to generate the design and coordinate subagents for task execution.