AI Coding Assistant
with Specialized Agents

Autonomous execution with validation. Analyzer → Planner → Editor → Validator. File validation prevents mistakes. Success criteria with auto-retry. $0-5/month vs $20-30/month subscriptions.

Get Started View on GitHub

Autonomous Execution

chu do "task" orchestrates 4 specialized agents: Analyzer → Planner → Editor → Validator. Auto-retry with model switching when validation fails.

Validation & Safety

File validation prevents creating unintended files. Success criteria auto-verified. No surprise scripts or configs. Supervised vs autonomous modes.

Intelligent Context

Dependency graph + PageRank identifies relevant files. 5x token reduction (100k → 20k). ML intent classification (1ms routing).

Radically Affordable

$0-5/month vs $20-30/month subscriptions. Use Groq for speed, Ollama for free. Auto-selects best models per agent.

Interactive Modes

chu chat for conversations. chu run for tasks with follow-up. Context-aware from CLI or Neovim plugin.

Manual Workflow

Break down complex tasks: chu research → chu plan → chu implement. Full control when you need it.

How It Works

Orchestrated agents with validation and auto-retry

graph TB User["chu do 'add feature'"] --> Orchestrator{Orchestrator} Orchestrator --> Analyzer["Analyzer
Understands codebase
Reads relevant files"] Analyzer --> Planner["Planner
Creates minimal plan
Lists files to modify"] Planner --> Validation["File Validation
Extracts allowed files
Blocks extras"] Validation --> Editor["Editor
Executes changes
ONLY planned files"] Editor --> Validator["Validator
Checks success criteria
Validates results"] Validator -->|Success| Done["Task Complete"] Validator -->|Fail| Retry["Auto-retry
with feedback"] Retry --> Editor style Analyzer fill:#3b82f6,color:#fff style Planner fill:#8b5cf6,color:#fff style Editor fill:#10b981,color:#fff style Validator fill:#f59e0b,color:#fff style Validation fill:#ef4444,color:#fff

Structured Workflow: Research → Plan → Implement

Go from feature idea to tested code with AI assistance at each step

Research

Understand your codebase before making changes

chu research "How does authentication work?"

Chuchu searches semantically, reads relevant files, and documents findings in ~/.chuchu/research/

Plan

Create detailed implementation plan with phases

chu plan "Add password reset feature"

Generates step-by-step plan with clear goals, file changes, and test requirements

Implement

Execute plan interactively or autonomously

chu implement plan.md
chu implement plan.md --auto

Interactive mode for control, autonomous mode for speed with automatic verification and retry

Complete Workflow Guide Read Tutorial

Quick Start

1. Install CLI

go install github.com/jadercorrea/chuchu/cmd/chu@latest
chu setup

2. Add Neovim Plugin

-- lazy.nvim
{
  dir = "~/workspace/chuchu/neovim",
  config = function()
    require("chuchu").setup()
  end,
  keys = {
    { "<C-d>", "<cmd>ChuchuChat<cr>", desc = "Toggle Chat" },
    { "<C-m>", "<cmd>ChuchuModels<cr>", desc = "Profiles" },
  }
}

3. Start Coding

chu chat "add user authentication with JWT"
chu research "best practices for error handling"
chu plan "implement rate limiting"

Features

Profile Management

Switch between model configurations instantly. Budget profile with Groq Llama 3.1 8B. Quality profile with GPT-4. Local profile with Ollama.

Cost Optimization

ML Intelligence

Context Optimization

Neovim Features

Why Chuchu?

Most AI coding assistants lock you into expensive subscriptions ($20-30/month) with black-box model selection and no validation. Chuchu gives you:

Read the full story: Why Chuchu?

Documentation