# AgentLed
> The automation engine built for AI agents.

AgentLed is an AI-native workflow automation platform where AI agents own, manage, and improve business workflows. It provides integrations, persistent business memory, approval controls, and access from Claude Code, Codex, Cursor, or any MCP client.

## Quick start

    npx @agentled/cli setup

The setup command opens browser authentication, selects a workspace, scaffolds a workspace folder, connects the MCP server, installs the AgentLed skill, and checks the company profile.

## Core pages

- [Home](https://www.agentled.ai/en): Product overview
- [What we solve](https://www.agentled.ai/en/what-we-solve): Business use cases
- [Pricing](https://www.agentled.ai/en/pricing): Plans and credits
- [Developers](https://www.agentled.ai/en/developers): MCP server, CLI, and API
- [Blog](https://www.agentled.ai/en/blog): Guides and product articles
- [AI VC Analyst](https://vc.agentled.ai): AI deal sourcing and investment-team support
- [Developer partners](https://www.agentled.ai/en/developer-partners): Build repeatable agent packages

## Documentation

- [Quick start](https://www.agentled.ai/en/docs/quick-start)
- [CLI guide](https://www.agentled.ai/docs/cli.md)
- [MCP guide](https://www.agentled.ai/docs/mcp.md)
- [Agentic operations patterns](https://github.com/Agentled/agentic-ops)
- [MCP server source](https://github.com/agentled/mcp-server)

## Core concepts

- Persistent knowledge graph memory across executions
- Human-controlled approval policies and observable execution history
- One connection surface for multiple providers and AI models
- White-label delivery for agencies and implementation partners
- MCP and CLI access for coding agents

## Contact and policies

- [Contact](https://www.agentled.ai/en/contact-us)
- [Privacy](https://www.agentled.ai/en/privacy-policy)
- [Terms](https://www.agentled.ai/en/terms-of-use)
- [AI disclosure](https://www.agentled.ai/en/ai-disclosure)
- [Sitemap](https://www.agentled.ai/sitemap.xml)