開発ツール
DeepSeek-Reasonix
A terminal coding agent built for DeepSeek models — reliable, cache-aware and cheap to run
★ 35.7k スターMIT
ガイドを見る →plandexは、plandex-ai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: plandex-ai/plandex）。GitHubスターは15,702、フォークは1,176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2025年10月3日です。
Open-source terminal AI coding agent built for large projects and multi-step, real-world tasks.
Decide what you want built or changed; Plandex suits multi-step work touching many files, and its chat mode can help flesh out the plan first.
Install Plandex with the one-line installer (on Windows, use WSL), which adds the plandex command with no dependencies.
Add your OpenRouter or other provider API key, or connect a Claude Pro/Max subscription when prompted on first run.
cd into your project directory, run plandex, describe the task, then review the sandboxed diff before applying changes.
Handles up to 2M tokens of context directly (about 100k per file) and can index directories of 20M tokens or more using tree-sitter project maps.
AI-generated changes stay in a cumulative diff sandbox, separate from your project files, until you review and apply them, with rollback for debugging.
Go from full-auto mode - loading files, implementing, executing commands and auto-debugging - to step-by-step review when you want control.
Combine models from Anthropic, OpenAI, Google and open-source providers, with curated packs trading off capability, cost and speed, plus context caching.
Every plan update is versioned, with branches to explore multiple paths or compare models, plus Git integration and commit-message generation.
Automatically debugs terminal commands such as builds, linters and tests, and can debug browser apps when Chrome is installed.
Clone the Plandex source if you want to build or inspect it; most users install the CLI directly instead.
git clone https://github.com/plandex-ai/plandex.git
cd plandexInstall Plandex with the official one-line, zero-dependency installer from the README.
curl -sL https://plandex.ai/install.sh | bashMove into your project directory and launch the REPL (pdx for short), then describe your task.
plandexPlandex is developed by plandex-ai and open-source contributors.
The CLI is open source under MIT. You run it in local mode with Docker and your own model-provider keys; the README notes Plandex Cloud is winding down and no longer accepts new users.
Yes, in practice: you supply your own OpenRouter or other provider key, or connect a Claude Pro/Max subscription, and pay that provider for model usage.
Plandex is a terminal-first agent focused on very large projects and multi-step plans, with a diff-review sandbox and plan branching; editor-based tools focus more on inline editing inside an IDE.
Run the one-line installer, cd into a project, type plandex, connect a provider, and describe the task.
A terminal coding agent built for DeepSeek models — reliable, cache-aware and cheap to run
The terminal runtime your coding agents live on — a Rust multiplexer built for agent sessions
Lightweight open-source AI coding agent built for Chinese developers.
Terminal-native AI coding assistant for real repositories.
OpenAI's lightweight coding agent that runs locally in your terminal
Spettro is a terminal coding assistant built in Go.
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 — plandex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plandex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。
Bridge local AI coding agents (Claude Code, Cursor, Gemini CLI, Codex) to messaging platforms (Feishu/Lark, DingTalk, Slack, Telegram, Discord, LINE, WeChat …
Convert documentation sites, GitHub repos and PDFs into Claude AI skills and RAG-ready knowledge assets.
Fine-tune and deploy 600+ LLMs and 300+ multimodal models with PEFT or full-parameter training.
Agentic IDE that runs 100+ parallel coding agents in isolated workspaces on your own subscriptions.
Open, event-based protocol that connects AI agents to user-facing apps with streaming and shared state.
Turn domain documents into high-quality LLM fine-tuning, RAG and evaluation datasets through a visual UI.