What problem does it solve? Producing a complete learning package for one chapter of a fluid mechanics textbook (analysis, curated coverage, tested Python functions, a teaching notebook, and 4-5 interactive HTML explainers) requires coordinating many specialized agents, quality gates, and publish steps, which is error-prone to do by hand. ## Core Features & Use Cases - Phased autonomous pipeline: Runs analyze, curate, design, implement, verify, review, viz, notebook, knowledge, and publish phases with gates, fix loops, and a commit per phase. - Parallel subagent orchestration: Launches design and implementation together, then review, per-explainer viz builders, and the notebook builder concurrently, with merge gates and retry limits. - Resumable progress tracking: Tracks phase status in progress.json and supports --from, --only, --consult, and --no-push flags for partial or supervised runs. - Use Case: Run /do-chapter 7 to turn Chapter 7 (Gravity Waves) of Kundu's Fluid Mechanics into verified fluidpy functions, an executed teaching notebook, five interactive explainers, and a published GitHub Pages update. ## Quick Start Ask the assistant to run /do-chapter 7 to execute the full chapter pipeline, or use /do-chapter 7 --from viz to resume from the explainer-building phase.
