cuga-agent
CUGA is an open-source generalist agent harness for the enterprise, supporting complex task execution on web and APIs, OpenAPI/MCP integrations, composable architecture, reasoning modes, and policy-aware features.
CUGA (Configurable Generalist Agent) is an open-source enterprise agent harness that executes complex tasks across web interfaces and APIs with built-in support for OpenAPI and MCP server integrations. It provides a production-ready agent framework with composable architecture, multiple reasoning modes (fast/balanced/accurate), policy-aware features for governance, and hybrid execution capabilities combining browser automation with API calls. Developers can configure domain-specific tools, policies, and workflows without rebuilding orchestration from scratch, achieving state-of-the-art performance on AppWorld and WebArena benchmarks.
Key Features
Use Cases
- 01Building enterprise agents that combine CRM API access with web form automation in hybrid workflows
- 02Creating governed AI assistants with human-in-the-loop approval gates and compliance policies
- 03Deploying multi-agent systems where specialized sub-agents handle CRM, email, analytics, and other domains
- 04Implementing document Q&A systems with agent-level and session-scoped knowledge bases
- 05Prototyping agentic workflows with visual design in Langflow integration
- 06Running benchmark evaluations on AppWorld and WebArena with reproducible agent configurations
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
cuga-agent — FAQ
What is CUGA?+
CUGA is an open-source generalist agent harness for enterprises that supports task execution on web and APIs, with built-in MCP server integration, OpenAPI support, multi-agent orchestration, and policy-based governance features.
How do I install CUGA?+
Clone the GitHub repository, create a Python 3.12+ virtual environment with uv, run 'uv sync' to install dependencies, create a .env file with your LLM API key (OpenAI, WatsonX, Azure, Groq, etc.), then start with 'cuga start demo'.
Which AI clients work with CUGA?+
CUGA is a standalone agent framework that integrates with LLM providers (OpenAI, WatsonX, Azure OpenAI, Groq, OpenRouter) rather than AI client applications. It can be exposed as a tool to other agents or deployed via its web UI.
Do I need API keys to use CUGA?+
Yes, you need an API key for at least one LLM provider (OpenAI, IBM WatsonX, Azure OpenAI, Groq, or compatible services). Some features like E2B cloud sandboxes or specific MCP servers may require additional keys.
How do I add MCP servers to CUGA?+
Edit src/cuga/backend/tools_env/registry/config/mcp_servers.yaml to add MCP server configurations with connection details, or use the manager UI ('cuga start manager') to configure MCP servers in draft mode and publish them to production.
Is CUGA free to use?+
CUGA itself is open-source and free, but you'll incur costs from your chosen LLM provider (OpenAI, WatsonX, etc.) and optional services like E2B cloud sandboxes.
How do I install cuga-agent?+
Open the source repository on GitHub and follow its README. cuga-agent is a agent — MCP Agents Market links you directly to the official repo.
Is cuga-agent free?+
cuga-agent is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.