About Runplane

What Runplane Does

Runplane is the execution control layer for AI agents. We sit between your agent and the real world. When an agent attempts to execute a tool, Runplane evaluates the action against your policies and enforces a decision: ALLOW, BLOCK, or REQUIRE_APPROVAL.

No action reaches your systems without passing through Runplane.

How It Works

The integration model is SDK-first:

1. Bring your existing agent tools from any framework
2. Runplane automatically maps tools into canonical action types
3. Baseline policies applied out of the box—customize as needed
4. Wrap tool execution with runplane.guard()
5. Decisions enforced at runtime before execution

You remain in full control. Policies are customizable for any action type.

Why This Matters

AI agents are moving from recommendations to execution. They trigger infrastructure changes, database mutations, financial operations, and communication workflows autonomously. Without runtime control, these actions execute without oversight.

Runplane provides the control layer that lets you deploy agents with confidence. Every action is evaluated. Every decision is logged. High-risk operations require human approval.

Who It's For

Runplane is built for engineering teams running agents in production:

• Platform teams deploying autonomous AI systems
• Security architects requiring governance over agent execution
• Reliability engineers managing blast radius and containment
• Compliance teams needing auditable records of AI decisions

Our Approach

We believe runtime control should be clear, operational, and precise. We avoid hype and focus on system behavior. Our documentation describes what Runplane does, not what it might do.

Bring your tools. Actions are mapped automatically. Baseline policies apply. Customize as needed. Wrap with guard(). Control execution.

Contact

All inquiries: support@runplane.ai