開発ツール
repowise
Repowise — codebase intelligence: health scores, auto-docs, git analytics and dead-code detection via MCP.
★ 7.2k スターAGPL-3.0
ガイドを見る →fallowは、fallow-rsが公開しているオープンソースの開発ツールプロジェクトです（GitHub: fallow-rs/fallow）。GitHubスターは5,022、フォークは171で、MITライセンスの下で公開されています。GitHubでの最終更新は2026年10月8日です。
Codebase intelligence for TypeScript/JavaScript — health, duplication, dead code and architecture.
Pick your pain: dead code, growing complexity, duplication, or architecture drift in a TS/JS repo.
Run npx fallow in your repository for a combined health, duplication and unused-code report.
Add it as a dev dependency, install git hooks, or add the GitHub Action in report-only mode first.
Audit only what your change introduces, preview auto-fixes with fallow fix --dry-run, and prove exports unused before deleting them.
Health scores, complexity hotspots, duplication, circular dependencies, architecture boundaries, design-system drift and unused code from a single codebase graph.
fallow audit reports only the findings your change introduces — ideal as a pull-request gate that never blames you for legacy issues.
fallow fix removes fixable issues (dry-run first), and dead-code --trace proves an export is unused before you delete it.
Ships as a CLI, GitHub Action, GitLab CI support, an LSP, an MCP server and a VS Code extension.
The analyzer contains no AI: the same input always gives the same output with stable fingerprints, and rule severities (error/warn/off) are configurable.
fallow agent install wires detected agents (Claude Code, Codex, Cursor) with an AGENTS.md task map, skills, the MCP server and commit/push gates.
Clone Fallow if you want the source; most users run it via npx.
git clone https://github.com/fallow-rs/fallow.git
cd fallowRun the full health, duplication and unused-code pass in any TypeScript/JavaScript repo.
npx fallowAudit just your recent changes, then preview the available auto-fixes.
npx fallow audit --base HEAD~15
npx fallow fix --dry-runFallow is maintained by fallow-rs, with documentation at docs.fallow.tools.
Yes — it is open source under the MIT license (the README describes it as Rust, MIT licensed).
No. Fallow is a local static analyzer with no AI inside; it runs entirely on your machine or CI.
ESLint lints style and local rules and knip focuses on unused files/exports; Fallow builds one graph across the codebase to combine health, duplication, architecture, drift and dead-code analysis with change-scoped audits.
Run npx fallow in your repo, read the report, then adopt the GitHub Action in report-only mode.
Repowise — codebase intelligence: health scores, auto-docs, git analytics and dead-code detection via MCP.
Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked con…
Compiler-backed AI coding agent that maps your codebase — calls, data flow, control flow, class hierarchy — before reading code.
A codebase knowledge graph that makes Claude Code, Cursor, Codex and Gemini faster, cheaper and repo-aware.
Enterprise-grade codebase intelligence for AI agents — hybrid semantic search over 40M+ LOC, locally.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
LibreAIカタログに掲載された開発ツールプロジェクトの中で最もスターの多いもの — fallowが合わなかった場合に検討したい、現実的な候補リストです。
The open source coding agent
The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
OpenAI's lightweight coding agent that runs locally in your terminal
カタログの中でfallowとスター数が近い開発ツールプロジェクト — 前後の棚もそのまま見て回れます。
Call MCP servers from TypeScript or the CLI — MCP tools disguised as a simple TypeScript API.
An OpenRouter-style registry for agent tools: one token, thousands of priced endpoints, your own keys too
An end-to-end framework for building custom AI applications and agents on top of your database.
Generates agent-first Go CLIs (plus MCP servers) for any API or website — with local SQLite sync built in
Local-first web search, fetch, crawl and research for AI coding agents over MCP — no API keys.
The official C# SDK for the Model Context Protocol — build MCP servers and clients in .NET