What problem does it solve? Understanding a large or unfamiliar codebase is slow and error-prone, and manually writing architecture documentation quickly becomes outdated. This Skill automates the generation of structured, module-level Wiki documentation for any code repository and keeps it synchronized through incremental updates. ## Core Features & Use Cases - Five-Phase Documentation Workflow: Analyzes the repository, clusters components into modules, generates per-module documents with Mermaid diagrams, produces a repository overview, and manages sessions. - Three-Tier Enhancement Modes: Automatically detects available MCP servers and selects deep enhancement (codebase-memory-mcp with Leiden clustering, Cypher queries, cross-service tracing), call-graph enhancement (CodeGraph), or standard CodeWiki mode. - LLM Wiki Knowledge Base: Query generated docs and archived notes with query_wiki, archive decisions and lessons with ingest_note, and check doc-code consistency with lint_wiki. - Incremental Updates: Detects changed files via git diff, maps them to affected modules, and regenerates only what changed instead of rewriting everything. - Use Case: Point the Skill at a monorepo and receive a complete Chinese-language Wiki with per-module architecture diagrams, complexity hotspots, dead-code detection, and a cross-service topology overview. ## Quick Start Generate Wiki documentation for the repository at the current path using the codewiki skill.
