Preloop
UnclaimedHuman approval and governance for AI agents, intercepting risky actions and ensuring critical decisions.
Visit WebsiteFreeVisit Website
TL;DR - Preloop
- Policy engine that intercepts risky AI agent actions and routes them for human approval
- Works with Claude Code, Cursor, and Windsurf via MCP — define what agents can do autonomously
- Sends approval requests via mobile app, Slack, and email for deployments, refunds, and data changes
Pricing: Free forever
Best for: Individuals & startups
Pros & Cons
Pros
- Adds a critical safety layer between AI agents and production systems
- Works with any MCP-compatible agent without modifying agent code
- Flexible policy engine lets teams fine-tune autonomy vs. oversight
- Multi-channel notifications ensure approvals are never missed
Cons
- Adds latency to agent workflows that require human approval
- Limited to MCP-compatible agents — not all AI tools support MCP yet
- New product with evolving policy configuration options
Key Features
Policy-based controls defining what AI agents can and cannot doHuman-in-the-loop approval for risky agent actionsMCP server integration for Claude Code, Cursor, and WindsurfNotifications via mobile app, Slack, Mattermost, and emailAudit log of all agent actions and approval decisionsCustom policy definitions per tool and action type
Pricing Plans
Free
Free
- MCP proxy for agent approvals
- Custom policy definitions
- Mobile, Slack, and email notifications
- Audit logging
About Preloop
By Toolradar Team·
Preloop is a policy engine and human approval layer for autonomous AI agents. It intercepts requests when AI agents attempt risky actions — like deployments, refunds, or data modifications — and routes them for human approval before execution. Teams define policies specifying what an AI agent can do autonomously, what it cannot do, and what requires approval. Works with any MCP-compatible agent including Claude Code, Cursor, and Windsurf, with notifications via mobile app, Slack, and email.
Reviews
No reviews yet. Be the first to review Preloop!
Explore More
Preloop FAQ
What is Preloop?
Preloop is a policy engine for AI agents that intercepts risky actions like deployments, refunds, or data changes and routes them to a human for approval before execution. You define what each agent can do autonomously and what needs oversight.
Which AI agents work with Preloop?
Preloop works with any MCP-compatible AI agent, including Claude Code, Cursor, and Windsurf. You connect your agent to Preloop via MCP and define policies for each tool the agent uses.
How do I get notified about approval requests?
Preloop sends notifications via its mobile app, Slack, Mattermost, or email when an AI agent attempts an action that requires human approval according to your defined policies.
Source: preloop.ai