The Governance Layer for Safe Agentic Automation

Control what AI agents can do, how much they can spend and what requires human approval. Audit every action with zero friction.

Start Free Trial Book a Demo https://calendar.app.google/FV95tXZtfGpPk7398

Preloop is an AI safety and control platform for teams deploying AI agents into real workflows. Use the preloop agents discover command to bring compatible local agent configs into Preloop, then apply policy enforcement, human approvals, model-gateway controls, budget visibility, runtime observability, and audit trails without rebuilding your stack. Explore how Preloop supports AI governance and AI Act readiness at /ai-act-readiness.

grid

Onboard existing agents into one control plane

Start with the preloop agents discover command to inspect compatible local setups, import supported tools and model metadata, and bring OpenClaw, OpenCode, Claude Code, Codex CLI, Gemini CLI, and MCP-compatible runtimes under control.

Govern tool access with policies and approvals

Define allow, deny, justification, and approval rules for MCP tools, parameters, and contexts. Protect deployments, databases, billing actions, secret access, and other high-impact workflows.

Route model traffic with budgets and attribution

Use the Preloop Gateway to control which models can be used, attribute usage to the right runtime, and keep model spend visible before costs run away.

Observe runtime sessions and collect governance evidence

Track what each agent attempted, which policy matched, who approved it, what the model traffic cost, and what happened next across sessions, tools, and workflows. Teams can use this operational evidence to support internal AI governance reviews and AI Act readiness work.

What is Preloop?

Preloop is an AI safety and control platform for teams deploying AI agents into real workflows. It combines governed tool access, human approvals, model-gateway controls, runtime observability, budget visibility, and audit trails in one control plane.

Who is Preloop for?

Preloop is built for engineering, platform, security, and operations teams adopting AI agents in real workflows. It is especially useful when agents can deploy code, access production data, change infrastructure, or spend money.

How does it work?

Connect your agent to Preloop over MCP, or onboard an existing local runtime with the preloop agents discover command. Preloop evaluates governed tool calls against your policies, and on supported managed paths it can route model traffic through the Preloop Gateway and tool access through the Preloop Tool Firewall.

Can I onboard agents I already use?

Yes. Preloop can discover compatible local configurations for OpenClaw, OpenCode, Claude Code, Codex CLI, Gemini CLI, and other MCP-compatible runtimes. On supported managed onboarding paths, it can import compatible tools and model metadata into your Preloop account.

Which AI agents does Preloop support?

Preloop works with OpenClaw, OpenCode, Claude Code, Codex CLI, Gemini CLI, Cursor, Windsurf, Cline, and other MCP-compatible agents and tools.

What actions can I control?

Any action exposed through MCP: deployments, database operations, secret access, cloud provisioning, billing changes, ticket automation, and internal tools. Policies can inspect arguments and context, not just tool names.

Do I need to modify my infrastructure or app code?

No. Preloop is designed to fit existing agent workflows without requiring SDKs or invasive infrastructure changes. Teams can add guardrails incrementally.

How do approval notifications work?

When a tool call needs approval, Preloop notifies the right people and waits. Approvers can review context, approve, reject, and leave guidance before the agent continues.

Will policies slow down my AI agents?

Only actions requiring approval pause for human input. Allowed actions run instantly. Denied actions fail immediately with a clear message. Most workflows run without any delay.

Do I get an audit trail?

Yes. Every action is logged with the attempted operation, inputs, policy decision, approver, timestamps, and final outcome. This supports security reviews, incident analysis, and compliance.

AI Act readiness with Preloop.">

Can Preloop help with EU AI Act readiness?

Yes. Preloop can help teams build operational controls and evidence for AI governance programs, including approval workflows, runtime visibility, policy enforcement, and audit trails. It is best positioned as part of an AI Act readiness program, not as a blanket legal compliance guarantee. See AI Act readiness with Preloop.

Does Preloop make my company automatically compliant with the EU AI Act?

No. Preloop does not replace legal interpretation, risk classification, conformity assessment, or broader compliance work. It helps teams implement technical controls, oversight workflows, and evidence collection that can support AI Act readiness and internal governance.

Onboard or connect your AI agent in minutes See AI Act readiness guidance /ai-act-readiness Start with preloop agents discover, or connect Preloop manually 
claude mcp add --transport http preloop https://preloop.ai/mcp/v1
  --header "Authorization: Bearer YOUR_PRELOOP_API_KEY"

# Create or edit ~/.cursor/mcp.json
{
  "mcpServers": {
    "preloop": {
      "transport": "http",
      "url": "https://preloop.ai/mcp/v1",
      "headers": {
        "Authorization": "Bearer YOUR_PRELOOP_API_KEY"
      }
    }
  }
}

{
  "mcpServers": {
    "preloop": {
        "transport": "http",
        "url": "https://preloop.ai/mcp/v1",
        "headers": {
            "Authorization": "Bearer YOUR_PRELOOP_API_KEY"
        }
    }
  }
}