Coding Agent
Roo Code
Open-source VS Code agent with role-specific coding modes
Our verdict
Free open-source VS Code agent with role-specific modes: Architect, Code, Debug, Test. Strong model flexibility. 23.7k GitHub stars. A focused Cline fork.
Best for
Developers who want Cline-style agentic coding with more structured role separation — Architect mode for planning, Code mode for implementation, Debug mode for fixing. Useful for complex tasks that benefit from keeping the AI's focus narrow.
Not for
Non-VS Code developers — Roo Code is VS Code only. Anyone wanting a managed hosted solution rather than BYOK.
Overview
Roo Code is a free, open-source AI coding agent for VS Code with 23.7k GitHub stars. Its defining feature is role-specific modes: Architect mode plans features without touching code, Code mode implements them, Debug mode traces errors, Test mode writes coverage. Each mode limits the AI's tool access to what's relevant — reducing runaway edits. Fully BYOK and model-agnostic.
What works
- +Role-specific modes (Architect, Code, Debug, Test) keep the AI focused on one job at a time
- +Fully free — no subscription, just API costs
- +Model-agnostic: works with Claude, GPT, Gemini, and local models
- +Permission-based command approval before any command runs
- +Open source — transparent about what it's doing and why
What doesn't
- −VS Code only — no JetBrains, no CLI-first workflow
- −Smaller community than Cline (23.7k vs 61k stars)
- −Mode switching adds cognitive overhead for simple tasks — sometimes you just want to ask and get an answer
- −Less enterprise support infrastructure than Cursor or Cline
What operators use it for
01
Planning Before Coding With Architect Mode
Before writing a single line, switch to Architect mode and ask Roo Code to plan the implementation. It reads the codebase, proposes an approach, and maps out the files it'll touch — all without editing anything. Switch to Code mode when the plan looks right. The separation prevents the common failure mode of AI agents that start editing before they understand the problem.
02
Isolated Debugging Sessions
Paste an error or point at a failing test. Debug mode narrows Roo Code's tool access to read-only analysis and targeted fixes — it doesn't refactor unrelated code while fixing your bug. Better surgical precision than a general-purpose agent that treats every task the same way.
03
Test Coverage on Demand
Point Test mode at a module with low coverage. It reads the implementation, identifies edge cases, and writes a suite covering them — without touching the implementation files. Useful before a refactor or release when you need confidence without adding risk.
Pricing
Free and open-source. BYOK — pay only for API calls to your chosen provider. No Roo Code subscription fee.
Continue exploring
Tool
AI Agent Cost Calculator
See what running Roo Code's typical workflow costs at your volume — pre-filled for this category.
Tool
The AI Agent Picker
Not sure Roo Code is right? Five questions, one recommendation.
Article
What Claude Skills actually are
Most builders think Skills are saved prompts. The architecture is different.
More in Coding agents
Claude Code
The next level up from the Claude you already use
Cursor
The AI-first IDE that 360k developers pay for
Aider
Open-source CLI pair programmer with model freedom
GitHub Copilot
The AI most developers already pay for, now with agent mode
Augment Code
The agentic coding tool built for large codebases
Amp
Sourcegraph's agentic coding tool with codebase context built in
OpenAI Codex
OpenAI's terminal-native coding agent
Cline
The open coding agent — 61k GitHub stars, 5M installs
Kilo Code
Open-source coding agent for VS Code, JetBrains, and CLI
OpenHands
The autonomous coding agent for full engineering tasks
Open dataset. This review is part of a structured dataset of every platform on the shortlist, published as platforms.json on GitHub under CC-BY-4.0.