maestro-learn

Guided codebase learning uses structured reading, evidence gathering, and pattern extraction to build reliable understanding. It walks through code paths section by section with forcing questions to create an understanding map.

Hypothesis-driven code investigation answers specific technical questions by collecting evidence, forming hypotheses, testing them, and writing a report. This ensures grounded understanding rather than superficial inspection.

Pattern decomposition extracts reusable design, behavioral, data, and error-handling patterns from a module or directory. It anchors findings and cross-references conventions to persist learning artifacts for repeatable analysis.

Second-opinion analysis reviews, challenges, and consults on code and plans from multiple perspectives. It applies to targeted engineering questions requiring grounded understanding and evidence-backed workflows.

Investigate the target code path through guided reading and evidence gathering to summarize key patterns, assumptions, and risks. This requires anchored findings and convention cross-references for reliable analysis.

Guided reading is needed when simple code reading is insufficient to build reliable understanding. It applies to code paths, modules, wiki entries, and technical decisions requiring evidence-backed, persisted learning artifacts.