Développement
Codewhale
Open-source Rust agent engine and terminal client with provider choice, approvals and receipts
★ 41,1 k étoilesMIT
Voir le guide →jcode est un projet de développement open source, créé par 1jehuang (1jehuang/jcode sur GitHub). Il compte 20 354 étoiles sur GitHub et 2 367 forks, il est publié sous la licence MIT et a été mis à jour pour la dernière fois sur GitHub le 8 octobre 2026.
A high-performance coding agent harness in Rust - fast TUI, swarms and every provider
Decide whether to use a subscription login (Claude etc.), an API provider, or a local server like Ollama or LM Studio.
Install it with the official installer for macOS/Linux or Windows.
Run jcode login, pick your provider or add an OpenAI-compatible endpoint or MCP servers in the config.
Open a project, choose a model, and give the agent its first coding task.
The README benchmarks fast startup (time to first frame/input) and low RAM, including headless swarm workers.
Keeps memory across sessions so the agent retains project context.
Run headless worker agents in parallel with configurable effort levels for bigger tasks.
Side panels, diagrams and info widgets render inside the TUI with smooth scrolling.
Built-in profiles for OpenRouter, DeepSeek, Kimi, Hugging Face and many other OpenAI-compatible endpoints, plus arbitrary custom endpoints.
Profiles for local runtimes such as Ollama and LM Studio, and self-hosted vLLM endpoints.
Get the source code from GitHub and enter the project directory.
git clone https://github.com/1jehuang/jcode.git
cd jcodeUse the installer for macOS and Linux (a PowerShell installer is provided for Windows 11).
# Run the install command from the README's Installation sectionAuthenticate with your provider, then start the agent in your project.
jcode login
jcodeIt is maintained by 1jehuang and contributors on GitHub.
Yes - jcode is MIT licensed and free. Models you connect may be paid subscriptions or APIs.
jcode itself charges nothing, but you need model access: a supported subscription login, your own API keys, or a free local model server.
Compared with Claude Code, jcode is provider-agnostic and Rust-built for speed and low memory, with swarm workers; compared with IDE agents, it is terminal-first.
Install with the README's installer, run jcode login, choose a model, and start a session in your project.
Open-source Rust agent engine and terminal client with provider choice, approvals and receipts
A native Rust port of the Pi coding agent: single binary, zero unsafe code, and hardened extensions.
Wake — every coding-agent session on your machine in one native app: find any conversation in seconds, read it in full, and resume it in your terminal
A tmux TUI for managing AI coding agents — status, prompts, worktrees, diffs.
A coding-agent harness optimized for smaller local LLMs — small models, serious coding work
The terminal runtime your coding agents live on — a Rust multiplexer built for agent sessions
Les autres projets de la catégorie Développement avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si jcode ne vous convient pas.
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
Projets de la catégorie Développement proches dans le catalogue, les plus proches en étoiles de jcode — continuez à explorer de part et d'autre.
Self-hosted TikTok and Douyin scraper API - no-watermark downloads, MCP and web console
Turn any GitHub repo into an interactive architecture diagram - just swap 'hub' for 'diagram'
Reverse engineer anything with agents - from app behavior down to native binaries
GenBI for AI agents - governed text-to-SQL and dashboards through an open context layer
Fine-tune huge models on small hardware with LoRA and other efficient methods
A high-performance AI pipeline engine - build LLM workflows as portable .pipe files
Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.