logo
AgentLed

cli.agentled.ai

Install AgentLed in your AI agent

One setup command connects your local agent client to AgentLed workflows, integrations, approvals, and memory.

Agent prompt

use cli.agentled.ai to install AgentLed

Works with Codex, Claude Code, OpenClaw, Hermes, and MCP-compatible clients.

Run setup

npx @agentled/cli setup

The setup opens authentication, creates or connects a workspace, and writes local MCP configuration.

Setup steps

  1. Install. Run the setup command from your project terminal.
  2. Authenticate. Connect or create your AgentLed workspace.
  3. Restart. Reload your AI client so MCP tools are available.

Try this first

Ask your agent to build a narrow workflow before publishing anything live.

Build me a lead-enrichment workflow:
- input: a company domain
- enrich via LinkedIn + Hunter
- score fit against my ICP
- save qualified leads to a Knowledge Graph list called "leads"

Use the AgentLed dry-run protocol before publishing.

Keep the first run in dry-run mode until you review outputs.

Instructions for your agent

After setup, restart the agent client and ask it to inspect available AgentLed tools.

  1. Run npx @agentled/cli setup from the project root.
  2. Restart the AI client after setup completes.
  3. Orient the agent:
    1. List available AgentLed tools.
    2. Inspect the workspace context.
    3. Use dry-run before publishing workflows.
  4. Relevant packages: @agentled/cli and @agentled/mcp-server.

Reference page: /en/cli. Headless docs: /docs/cli.md. Raw install file: /install.md.

Links