Secure the AI agents, apps and chats your team runs.

Guides and reference for the Beam CLI, dashboard, browser extension, App monitoring and MCP server — each one plain Markdown your coding agent can read too.

Start the quickstart Open the dashboard

Two commands to your first event

npm i -g @agent-beam/beam
beam setup

# No Node.js? Standalone installer
curl -fsSL https://agentbeam.com/install.sh | bash

Where do you want to start?

Pick the line that sounds like you. Three pages each, in the order to read them.

I run the workspace

Set Beam up for the whole company and see all of it.

I own security

Review what agents did and keep everyone within policy.

I ship an AI app

Know when the app is attacked, and what got through.

I code with agents

Connect Claude Code, Codex or Cursor and keep shipping.

I use AI in the browser

Use ChatGPT, Claude and Gemini without leaking data.

Browse by section

Named like the dashboard, so the page you need is where you'd look for the screen.

9 guides

Coding agents

See and steer what Claude Code, Codex, Cursor and others do on every laptop.

Connect coding assistantsLocal StudioHistorical extraction 4 guides

AI websites

Protect what people paste into ChatGPT, Claude and Gemini with the browser extension.

Browser extensionInstall the extensionReview, redact & allow once 15 guides

App monitoring

Watch the apps you ship for attacks, scraping and exposed secrets.

App monitoring and dependency checksApp monitoring, end to endAfter an attack: responding to a finding 7 guides

MCP & skills

Scan the skills, MCP servers and packages agents use, and share the good ones.

Scan skills & MCP configsScan npm & PyPI packagesTest skills, MCP servers and packages before agents use them 9 guides

Policies & controls

Decide what agents may do, review what they did, and get alerted.

Workspace policiesPolicy presetsDetection & custom rules 5 guides

Workspace & admin

People and roles, data handling, self-hosting and your plan.

Members & rolesWorkspace & account settingsData, redaction & export 3 guides

Cookbooks

Ordered, end-to-end recipes with a check after every step.

CookbooksAgent monitoring, end to endEvaluate agents before you trust them

App monitoring for your stack

Ordered recipes from nothing to fully set up, with a check after every step. Start with the App monitoring cookbook, or pick your framework.

Next.js Express & Fastify NestJS Django Flask FastAPI Ruby on Rails Laravel WordPress Spring Boot Go ASP.NET Core

Built for coding agents

Docs your agents can read.

Every page is plain Markdown at the same address plus .md, and llms.txt lists them all. Copy for agents on any page copies a prompt like this one, so Claude Code, Codex or Cursor can work through a guide with you.

How cookbooks work

Read the AgentBeam cookbook at https://docs.agentbeam.com/cookbooks/app-monitoring.md and work through it with me step by step for this project.

Built for enterprise

Least privilege, every change audited, and your data where you choose.

Self-hostRun the dashboard in your own cloud.Data & redactionWhat leaves a machine, and what never does.Least privilegeOwners, admins, security managers, auditors.Audit & exportEvery change recorded, exportable to CSV.SIEM & on-callSigned webhooks, PagerDuty, Slack and Teams.