This framework provides an AI-driven approach to software development, focusing on plan-first workflows with explicit user approval before execution. It targets developers and power users seeking to integrate AI assistance into their coding processes, offering benefits like automated quality checks, multi-language support, and adherence to custom coding standards, all orchestrated through the OpenCode CLI.

How It Works

OpenAgents employs a plan-first, approval-based execution model. User requests are analyzed by main agents (e.g., openagent, opencoder), which then propose an implementation plan for user approval. Upon approval, the agent executes the plan incrementally, delegating specialized tasks like coding, testing, and review to dedicated subagents. This approach ensures human oversight and integrates user-defined coding patterns from context files, enhancing control and quality.

Quick Start & Requirements

1. Prerequisite: Install the OpenCode CLI following its official guide.
2. Installation: Use the provided install.sh script via curl. Options include interactive installation or profile-based installs (e.g., developer, full, advanced).
   • Example (Developer profile): curl -fsSL https://raw.githubusercontent.com/darrenhinde/OpenAgents/main/install.sh | bash -s developer
3. Requirements: Bash 3.2+ is required.
4. Links:
   • OpenCode CLI Docs: https://opencode.ai/docs
   • Installer Script: https://raw.githubusercontent.com/darrenhinde/OpenAgents/main/install.sh

Highlighted Details

• Multi-language Support: Works with TypeScript, Python, Go, Rust, and more.
• Plan-First Workflow: Agents propose plans before implementation, requiring user approval.
• Incremental Execution: Tasks are executed step-by-step with built-in validation.
• Integrated Quality: Features automatic testing, type checking, and code review.
• Context-Aware: Agents load and follow user coding standards from .opencode/context/ files.
• System Builder: An interactive tool to generate custom AI architectures and systems.
• Optional Add-ons: Telegram notifications and Gemini AI image generation tools are available.

Maintenance & Community

Contributions are welcomed via a dedicated contributing guide. The project shows recent activity via its last commit badge. Community channels like Discord or Slack are not explicitly mentioned.

Licensing & Compatibility

The project is licensed under the MIT License, which is permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

The framework is currently optimized for the OpenCode CLI; support for other AI coding tools like Cursor or Claude Code is planned for future releases. The README notes that the repository has evolved since a demo video, with a video-simple branch available for users following older tutorials.