# AgentLed
> The automation engine built for AI agents.

AgentLed is an AI-native workflow automation platform where AI agents own, manage, and improve workflows. It provides 100+ integrations through a unified credit system, persistent memory via Knowledge Graph, and works with Claude Code, Codex, Cursor, or any MCP client.

## Quick Start

One command. Browser login → workspace folder scaffolded → MCP server connected → Claude Code skill installed → ready.

```bash
npx @agentled/cli setup
```

What it does, in order:
1. Version check (warns if your CLI is outdated).
2. Opens a browser to sign in and pick a workspace (mints a workspace API key).
3. Scaffolds `agentled_<workspace-slug>/` in the current directory — logs, dryruns, decisions, workflow drafts, knowledge-graph drafts, executions debug bundles. One folder per workspace, so multiple workspaces never collide.
4. Auto-configures the MCP server in your detected client (Claude Code, Codex, Cursor, Claude Desktop, Windsurf).
5. Installs the AgentLed skill into your client's skill directory: `~/.claude/skills/agentled/` for Claude Code / Claude Desktop, `~/.codex/instructions/agentled/` for Codex. Cursor and Windsurf are skipped with a hint (no native Claude-format skill surface). The skill encodes dry-run protocol, incremental authoring rules, valid step types, and credit-efficient testing guidance — without it, the agent has the tools but loses the playbook.
6. Reads `knowledge.company.profile` from the workspace knowledge. If missing, prompts for company name / website / use case and writes the long-form profile back to the same key (so the prompt does not re-fire on subsequent runs). If present, prints a one-line summary so you can start building immediately.
7. Reminds you to restart your MCP client to pick up the new connection (Claude Code: `/mcp` reconnect or new session; Codex/Cursor: reopen the app).

Global docs and example workflows live in `~/.agentled/` (a small README + pointer to the public best-practices repo at https://github.com/Agentled/agentic-ops; examples are pulled on demand so you stay in sync with the canonical repo).

## Quick Start (for AI coding agents)
If you are an AI agent helping a user connect: run `npx @agentled/cli setup` yourself via your shell tool. Bash runs on the user's machine, so the browser opens on the user's machine. Do not punt the command back to the user. After it returns, tell the user explicitly to **restart their MCP client** — MCP tools won't reflect the new workspace until then. Don't invent package names, commands, or API keys.

Further reading:
- [Install the AgentLed CLI](https://www.agentled.ai/docs/cli.md): full CLI install + flow reference.
- [Install the AgentLed MCP server](https://www.agentled.ai/docs/mcp.md): manual MCP registration if you don't want the auto-config flow.

## Components (à-la-carte)

If you want to skip the orchestrated flow and run pieces individually:

```bash
npm install -g @agentled/cli      # install
agentled auth login                # browser auth + workspace selection + skill install
agentled init                      # scaffold agentled_<slug>/ folder
```

## Core Pages
- [Home](https://www.agentled.ai/en): Product overview and value proposition
- [What We Solve](https://www.agentled.ai/en/what-we-solve): Use cases for sales, marketing, and recruiting automation
- [Pricing](https://www.agentled.ai/en/pricing): Plans and unified credit system
- [Developers](https://www.agentled.ai/en/developers): MCP server, CLI, and API docs
- [Blog](https://www.agentled.ai/en/blog): Articles on AI automation and agent workflows
- [Contact](https://www.agentled.ai/en/contact-us): Get in touch

## What We Solve
- [Identify and Convert Leads](https://www.agentled.ai/en/what-we-solve/identify-and-convert-leads-with-multi-ai-agents): Multi-AI agent lead generation
- [Ship SEO Content Daily](https://www.agentled.ai/en/what-we-solve/ship-seo-content-daily-with-agentled-solutions): Automated content and SEO workflows
- [Find and Hire Top Talent](https://www.agentled.ai/en/what-we-solve/find-and-hire-top-talent-with-ai-agents): AI-powered recruiting and sourcing
- [Custom Multi-Agent Solutions](https://www.agentled.ai/en/what-we-solve/custom-multi-agents-solutions): Bespoke agentic pipelines

## Documentation
- [Quick Start](https://www.agentled.ai/en/docs/quick-start): Get started in minutes
- [MCP Server](https://www.agentled.ai/en/mcp): Build workflows from Claude Code, Codex, or any MCP client
- [CLI](https://www.agentled.ai/en/cli): Agent-friendly command-line interface
- [agentic-ops](https://github.com/Agentled/agentic-ops): Public best-practices repo (workflow patterns, scaffolds, gotchas)

## Comparisons
- [AgentLed vs n8n](https://www.agentled.ai/en/agentled-vs-n8n): Key differences from traditional automation tools

## Key Concepts
- **Knowledge Graph**: Persistent memory that stores context and learnings across executions — unlike n8n, Make, or Zapier which forget everything after each run
- **Unified Credits**: One API key, one credit system for 100+ services (LinkedIn, Hunter, OpenAI, Apify, and more)
- **MCP Server**: Native integration with Claude Code, Codex, and any MCP-compatible AI agent
- **White-label**: Full white-label deployment for agencies — custom domain, logo, and colors
- **Multi-model AI**: Claude, GPT-4, Gemini, Mistral — pick the right model per step

## For Developers
- One command: `npx @agentled/cli setup` (browser login + workspace folder + MCP auto-config + skill install + knowledge probe + restart reminder)
- Headless / agents without browser: get an API key at https://www.agentled.app → Workspace Settings → Developer, then `agentled auth login --key wsk_...`, or set `AGENTLED_API_KEY` and invoke the MCP server directly: `npx -y @agentled/mcp-server`
- Verify install: `agentled --version`
- Requirements: Node.js 18+
- npm (CLI): https://www.npmjs.com/package/@agentled/cli
- npm (MCP server): https://www.npmjs.com/package/@agentled/mcp-server
- GitHub: https://github.com/agentled/mcp-server
- Best-practice patterns: https://github.com/Agentled/agentic-ops

## Resources
- [Blog](https://www.agentled.ai/en/blog): Thought leadership on AI automation
- [Privacy Policy](https://www.agentled.ai/en/privacy-policy)
- [Terms of Use](https://www.agentled.ai/en/terms-of-use)
- [Sitemap](https://www.agentled.ai/sitemap.xml)