Claude Code plugin that runs a multi-agent pipeline over any codebase or wiki, outputting an interactive knowledge graph with search, chat, diff impact, and guided tours.
Key Takeaways
Six specialized agents (project-scanner, file-analyzer, architecture-analyzer, tour-builder, graph-reviewer, domain-analyzer) run in parallel batches of 20-30 files with incremental update support.
Graph output is plain JSON stored in .understand-anything/; commit it once and teammates skip the pipeline entirely, useful for onboarding and PR reviews.
/understand-knowledge targets Karpathy-pattern LLM wikis: deterministic wikilink parser plus LLM agents extract entities, claims, and implicit relationships into a force-directed graph.
Works natively with Claude Code via plugin marketplace; Cursor and VS Code/Copilot auto-discover via config files; Codex, Gemini CLI, and others use AI-driven install instructions.
Persona-adaptive UI adjusts detail level for junior dev, PM, or power user roles; diff impact analysis shows ripple effects before committing.
Hacker News Comment Review
Both commenters questioned whether knowledge graph visualizations provide real utility beyond aesthetics, drawing comparisons to Obsidian’s graph view which is widely considered impressive-looking but low-value in practice.
No commenter provided benchmark evidence or workflow comparison against just using Claude Code or Codex directly, leaving the core value claim unverified.
Notable Comments
@analogpixel: “is this like Obsidian’s graph view? Looks pretty/makes cool screenshots but has no actual value”
@docheinestages: asks for evidence this beats raw Claude Code or Codex, noting a “gazillion” competing codebase-understanding graph tools already exist.
