summarize-architecture

To summarize an unfamiliar codebase into a one-page architecture brief, trace an end-to-end request through the code, identify state stores and external dependencies, and present the findings as a structured diagram-in-prose with explicit uncertainty for opaque regions.

Mapping system boundaries and state stores during codebase analysis involves tracing data flows to identify where state lives, distinguishing authoritative versus derived data, and highlighting trust or process boundaries that introduce risk within the architecture brief.

Preparing for a code review onboarding a new repository quickly requires capturing startup and shutdown sequencing, enforced conventions, and major component integrations into a single-page structured brief suitable for technical alignment.

Generating an architecture diagram directly from source code without external dependencies produces an ASCII-style component and data-flow explanation that clarifies request paths and responsibilities while explicitly marking opaque regions that lack clear visibility.

The limitations of a one-page architecture brief for complex systems include reduced detail granularity and the need to explicitly flag opaque regions where implementation visibility is limited, ensuring reviewers know where uncertainty remains.

Finding entry points before generating an architecture brief is a prerequisite step that enables the system to trace an end-to-end request accurately and map the lifecycle overview before summarizing the architecture.