AgentConsent: Auditable permissions layer for AI agents

A permission and compliance checkpoint for everything your AI agents do.

Problem

Tools like Breadcrumb record everything, Call4me can cancel subscriptions or book appointments, and multi-agent systems share tools and state, risking overreach or duplicate actions. Users increasingly worry about AI agents acting on sensitive operations without consent, yet there's no simple way to define what an agent may or may not do and keep an audit log.

Solution

A middleware/agent wrapper that lets users define fine-grained permissions per AI agent (e.g., read email only, no sending, payment requires confirmation), logs all actions to a verifiable audit trail, and supports one-click revocation. Start as a macOS menubar tool with optional API.

Users

Power users and small teams delegating personal or workflow tasks to AI agents, especially involving money, email, or appointments.

MVP

Define simple permission rules based on URL, app name, and action type; intercept and log sensitive agent actions; provide an audit log viewer and one-click revoke; export logs as JSON for compliance; macOS menubar quick toggle.

Revenue

One-time personal license ($19–$39), team subscription ($5/user/month), or API billed per audit event.

Why now

AI agents are moving from chat to execution, but the permission and safety layer is nearly empty; big labs focus on models, app startups on use cases, leaving a niche for an indie to define.