- Category
- Developer Tools
- Rank
- No. 1658Tools index
- Pricing
- Open Source
- Type
- TOOL
- Builder
- abhigyanpatwari
- GitHub
- 16 stars
- Date
About
Interactive knowledge graph for codebases — a vibe-coder companion for exploring and reasoning about source code.
What it does
Prowl is a desktop companion for watching AI-driven code changes. Open a folder, ZIP archive, or public GitHub repository, and it maps files, symbols, and relationships. Its graph updates as local files change. Search, chat, code viewing, editing, and a terminal sit alongside the visualization.
Why it's ranked here
Prowl has a coherent job: reduce the effort of understanding what an AI coding agent touched and what depends on it. The live watcher, change-impact analysis, semantic search, and MCP access reinforce that job. The repository reports large token reductions, although its published benchmark covers one 40-file project and mixes measurements with estimates.
What's good
It combines visual inspection with agent access to the same code model. The graph distinguishes containment, definitions, imports, calls, inheritance, and implementations. Cached snapshots support faster reopening, while incremental updates preserve the existing map. Local Ollama support offers chat without cloud model costs, and MCP can return compact architectural context.
Tradeoffs
The project labels itself beta. macOS builds are not code-signed, so first launch requires bypassing quarantine or building from source. Built-in chat needs a configured provider, and cloud providers require API keys. The headline token savings come from a small published trial, with several comparison figures estimated. Language parsing is limited to the documented supported set.
How to use it well
Use Prowl beside an AI coding agent when repository orientation, dependency tracing, and change monitoring consume repeated context. Start with the graph, inspect hotspots, then let the agent query focused context through MCP. It best suits visual learners and teams handling unfamiliar code. Treat impact results as navigation guidance, not proof of correctness or a replacement for tests.
Technical notes+
The Electron and React entry flow in src/App.tsx handles ZIPs, cloned repositories, native folders, cached snapshots, incremental refreshes, embeddings, and filesystem watching. src/hooks/useSigma.ts renders Graphology data with Sigma, ForceAtlas2, overlap cleanup, highlighting, and change animations; src/lib/graph-adapter.ts converts code nodes and relationships into display attributes. src/mcp/types.ts defines 13 internal tool names, including chat history and status, while the README presents 12 user-facing MCP tools. src/services/zip.ts strips dominant archive roots and applies ignore rules. package.json builds an Electron application plus a bundled Node 18 ESM MCP server and packages macOS, Windows, and Linux targets.
Observed
- License
- Business Source License 1.0 (BSL-1.0).
- Primary language
- TypeScript, with React used for the desktop interface.
- Application stack
- Electron, React, TypeScript, Vite, Tailwind CSS, Sigma, and Graphology.
- Install surface
- Release artifacts include macOS DMGs, a Windows installer, and Linux AppImage or Debian packages; source builds use npm.
- Interfaces
- Electron desktop application with graph, terminal, chat, search, code viewer, editor, and a stdio MCP server.
- Platform support
- macOS on Apple Silicon and Intel, Windows x64, and Linux x86_64.
- Indexed source languages
- JavaScript, TypeScript, Python, Java, Go, Rust, C, C++, and C#.
- Runtime prerequisite
- Source builds and the bundled MCP server target Node.js 18 or newer.
Read from README.md, package.json, src/App.tsx, src/main.tsx, src/vite-env.d.ts, src/lib/utils.ts, src/mcp/types.ts, src/lib/louvain.ts, src/services/zip.ts, src/lib/constants.ts, src/hooks/useSigma.ts, src/lib/file-icons.ts, src/types/pipeline.ts, src/hooks/useSettings.ts, src/lib/graph-adapter.ts.
What it can do
Generate interactive knowledge graph of codebase
Source code repository → Visual knowledge graph showing code relationships
Explore code dependencies and connections
Codebase components or files → Interactive visualization of code relationships and dependencies
Navigate code structure visually
Source code files and directories → Interactive map of code organization and structure
Analyze code relationships and patterns
Source code repository → Insights about code connections and architectural patterns
Search and discover code components
Search queries or code elements → Relevant code components and their relationships
Reason about code architecture
Codebase structure and components → Analysis and insights about code design and architecture
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.