> ## Documentation Index
> Fetch the complete documentation index at: https://docs.permissionslip.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

> Human approval for AI agent actions

# Permission Slip

Permission Slip gives you a human-in-the-loop approval layer for your AI agents. When an agent wants to take a sensitive action — sending an email, calling an API, modifying data — Permission Slip pauses execution and asks you to approve or deny the request before anything happens.

## Why Permission Slip?

AI agents are getting more capable, but "capable" and "trusted to act unsupervised" are different things. Permission Slip bridges that gap:

* **You stay in control.** Every action an agent wants to take can require your explicit approval before it executes.
* **You see what's happening.** A real-time dashboard shows pending requests, recent activity, and which agents are active.
* **You set the rules.** Standing approvals let you pre-authorize routine actions so you're only interrupted for things that matter.
* **You get notified instantly.** Email, SMS, and browser push notifications make sure you never miss a request.

## Core Concepts

<CardGroup cols={2}>
  <Card title="Agents" icon="robot" href="/agents/overview">
    AI systems that connect to Permission Slip and submit actions for approval.
  </Card>

  <Card title="Connectors" icon="plug" href="/connectors/overview">
    Integrations that define what actions an agent can take — email, APIs, databases, and more.
  </Card>

  <Card title="Approvals" icon="check" href="/approvals/overview">
    The review process where you approve or deny each action request from your agents.
  </Card>

  <Card title="Standing Approvals" icon="shield-check" href="/approvals/standing-approvals">
    Pre-authorizations that let agents repeat trusted actions without asking every time.
  </Card>
</CardGroup>

## How It Fits Together

```
Agent wants to act → Permission Slip intercepts → You approve or deny → Action executes (or doesn't)
```

1. You **register an agent** and give it credentials to connect to Permission Slip.
2. You **enable connectors** that define what the agent is allowed to request.
3. When the agent wants to act, it **submits a request** through the Permission Slip API.
4. You **review and approve** (or deny) the request from your dashboard or via notification.
5. If approved, Permission Slip **executes the action** immediately via the connector and the agent receives the **result** — or, if denied, the agent is notified and stops.

Ready to get started? [Create your account →](/getting-started/create-account)

<Info>
  **Building an agent?** Skip straight to the [Agent Integration Guide](/guides/agent-integration) for the full API walkthrough — registration, authentication, approval flow, and action execution.
</Info>