pipe

To automate a multi-skill workflow pipeline, you can orchestrate sequential execution by splitting a task into phases and routing each phase to the appropriate skill. This coordinates methodologies like TRIZ, tracing, and TDD into a cohesive pipeline.

Combining analysis and implementation skills requires a meta-skill orchestrator that propagates results phase-to-phase. This ensures context continuity, allowing you to sequence tasks from problem statement analysis directly to working code generation.

Yes, you can chain multiple agents together to run phase by phase. The orchestrator prepares per-phase prompts and executes agents sequentially, feeding the output of each phase into the next to produce a comprehensive final report.

Pipeline orchestration handles context propagation by reading each skill's SKILL.md, preparing tailored prompts, and passing results directly from one phase to the next. This ensures full continuity across varying methodologies like UI changes and TDD workflows.

No, you do not need to manually prepare prompts for each phase. The orchestration mechanism automatically reads the required skill documentation and generates the necessary per-phase prompts to run the agents sequentially.

You should use a multi-skill pipeline when a task requires combining multiple methodologies, such as TRIZ analysis followed by TDD implementation. Running them separately breaks context continuity, whereas a pipeline propagates results end-to-end.