Trae Agent

Stale
GitHub Python MIT

Description

Trae Agent is an LLM-based agent for general purpose software engineering tasks.

Key Features

  • LLM-based general-purpose software engineering agent with CLI interface for natural language instructions
  • Multi-provider support: OpenAI, Anthropic, Google Gemini, OpenRouter, Doubao, and Ollama
  • Lakeview feature providing concise step-by-step summarization of agent actions
  • Rich tool ecosystem including file editing, bash execution, sequential thinking, and task completion
  • YAML-based configuration with environment variable fallback for flexible deployment
  • Docker mode for isolated agent execution in containers with custom images and Dockerfiles

Use Cases

💡 Automated software engineering tasks from natural language bug reports and feature requests
💡 Repository-level code modification, testing, and debugging workflows
💡 Research on AI agent architectures with transparent and modular codebase
💡 Containerized code execution for security-isolated development environments
💡 Iterative development through interactive conversational coding sessions

Quick Start

git clone https://github.com/bytedance/trae-agent.git && cd trae-agent && uv sync --all-extras && source .venv/bin/activate && trae-cli run "Create a hello world Python script"

Related Projects