Modo

Name: Modo
Availability: OnlineOnly
Author: mohshomis

An open-source AI desktop IDE that plans before it codes, turning prompts into structured requirements, design, and tasks before generating code.

Visit Website

At a Glance

Pricing

Open Source

Fully free and open-source under the MIT license. Clone, build, and run locally.

Engagement

Available On

Windows

macOS

Linux

API

VS Code

mohshomismohshomis builds Modo, an open-source AI desktop IDE that in…

Listed Apr 2026

About Modo

Modo is an open-source AI-powered desktop IDE built on top of Void (a VS Code fork) that introduces spec-driven development — transforming prompts into structured plans before writing a single line of code. It guides developers through a requirements → design → tasks → code workflow, keeping context persistent across sessions. Licensed under MIT, Modo is community-maintained and freely hackable.

Spec-Driven Development: Create a spec with Cmd+Shift+S; Modo generates requirements.md, design.md, and tasks.md files in .modo/specs/ and executes tasks one by one.
Task CodeLens: Inline "▶ Run Task" buttons appear directly in tasks.md, with status indicators (spinner, checkmark) as the agent works through each item.
Steering Files: Markdown rules in .modo/steering/ are automatically injected into every AI interaction, supporting always, fileMatch, and manual inclusion modes.
Agent Hooks: JSON configs in .modo/hooks/ automate actions around 10 agent lifecycle events (e.g., fileEdited, promptSubmit, preToolUse), supporting askAgent or runCommand actions.
Autopilot / Supervised Toggle: A status bar pill switches between fully autonomous agent execution and a supervised mode that pauses for user approval.
Parallel Chat Sessions: Multiple chat sessions run as tabs, each with its own thread, context, and history, plus a searchable history panel.
Subagents: Spawn parallel agents for independent subtasks, each with its own thread and tool access.
Powers: Installable knowledge packages bundling documentation, steering files, and MCP configs for TypeScript, React, Testing, API Design, Docker, and custom URLs.
Vibe and Spec Modes: Choose free-form exploration (Vibe) or structured spec-driven development (Spec) at session start.
Multi-Provider LLM Support: Inherited from Void — supports Anthropic, OpenAI, Gemini, Ollama, Mistral, Groq, and OpenRouter out of the box.
MCP Integration: Full Model Context Protocol support for tool use and external integrations.
Quick Start: Clone the repo, run npm install (Node 20 required), npm run buildreact, npm run watch, then ./scripts/code.sh to launch.

Community Discussions

Be the first to start a conversation about Modo

Share your experience with Modo, ask questions, or help others learn from your insights.

Pricing

OPEN SOURCE

Open Source (Free)

Fully free and open-source under the MIT license. Clone, build, and run locally.

Spec-driven development
Task CodeLens
Steering files
Agent Hooks
Autopilot / Supervised toggle

Capabilities

Key Features

Spec-driven development (requirements → design → tasks → code)
Task CodeLens with inline Run Task buttons
Steering files for persistent project rules
Agent Hooks with 10 lifecycle event types
Autopilot / Supervised toggle
Parallel chat sessions as tabs
Subagents for parallel task execution
Powers (installable knowledge packages)
Vibe and Spec session modes
Multi-provider LLM support (Anthropic, OpenAI, Gemini, Ollama, Mistral, Groq, OpenRouter)
MCP integration
Slash commands
Workspace initialization
Export conversation as Markdown
Custom Modo Dark theme
Searchable conversation history

Integrations

Anthropic

OpenAI

Gemini

Ollama

Mistral

Groq

OpenRouter

MCP (Model Context Protocol)

VS Code extensions

API Available

View Docs

Back to all tools Suggest an edit

About Modo

Spec-Driven Development: Create a spec with Cmd+Shift+S; Modo generates requirements.md, design.md, and tasks.md files in .modo/specs/ and executes tasks one by one.
Task CodeLens: Inline "▶ Run Task" buttons appear directly in tasks.md, with status indicators (spinner, checkmark) as the agent works through each item.
Steering Files: Markdown rules in .modo/steering/ are automatically injected into every AI interaction, supporting always, fileMatch, and manual inclusion modes.
Agent Hooks: JSON configs in .modo/hooks/ automate actions around 10 agent lifecycle events (e.g., fileEdited, promptSubmit, preToolUse), supporting askAgent or runCommand actions.
Autopilot / Supervised Toggle: A status bar pill switches between fully autonomous agent execution and a supervised mode that pauses for user approval.
Parallel Chat Sessions: Multiple chat sessions run as tabs, each with its own thread, context, and history, plus a searchable history panel.
Subagents: Spawn parallel agents for independent subtasks, each with its own thread and tool access.
Powers: Installable knowledge packages bundling documentation, steering files, and MCP configs for TypeScript, React, Testing, API Design, Docker, and custom URLs.
Vibe and Spec Modes: Choose free-form exploration (Vibe) or structured spec-driven development (Spec) at session start.
Multi-Provider LLM Support: Inherited from Void — supports Anthropic, OpenAI, Gemini, Ollama, Mistral, Groq, and OpenRouter out of the box.
MCP Integration: Full Model Context Protocol support for tool use and external integrations.
Quick Start: Clone the repo, run npm install (Node 20 required), npm run buildreact, npm run watch, then ./scripts/code.sh to launch.

Modo