Building keel

productHunt

Momen Adel  • Peerlist

<h3>Control Plane for AI Agents</h3>Keel provides a safety layer for your AI agents, sitting between the agent and its tools to ensure secure and controlled actions. It intercepts every AI action, checking it before it runs.<ul><li>Safe actions run instantly.</li><li>Risky actions (like sending emails, deleting data, or moving money) require your approval.</li><li>Dangerous actions are blocked entirely.</li><li>Every decision is recorded in a tamper-proof log, allowing you to undo mistakes with a single click.</li></ul><h3>How it Works</h3>Keel acts as a proxy, routing your agent's actions through its control plane. This allows for real-time decision-making on each action based on its risk level.<ol><li>Connect: Point your agent to Keel with a simple configuration change – no code modifications needed.</li><li>Check: Keel reviews every action the moment the agent attempts it.</li><li>Decide: Safe actions proceed, risky actions wait for your OK, and dangerous actions are blocked.</li><li>Undo: Reverse any mistaken actions from the activity log.</li></ol><h3>Key Features</h3><ul><li>Brake: Prevents risky actions by requiring user approval.</li><li>Record: Maintains a permanent, tamper-proof log of all agent activities.</li><li>Undo: Enables one-click reversal of agent mistakes.</li><li>The Ledger: Provides a detailed, hash-chained record of every decision, with secrets redacted.</li></ul>Keel is designed to be compatible with any MCP agent and can be set up in minutes, offering a robust solution for managing and securing AI agent operations.