> This page is for Tools.

> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.moveworks.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.moveworks.com/_mcp/server.

# Assistants, Agents, & Plugins

# Assistants & AI Agents

**Moveworks Assistants** are agentic systems that can help you accomplish any task: writing emails, managing your calendar, submitting time off, creating expense reports, analyzing CRM data, and [more](https://marketplace.moveworks.com/).

**Plugins** are skills that you give to AI Assistants.

1. **Conversational Plugins** are triggered by the **[Agentic Reasoning Engine](/harness/agentic-reasoning-engine/)**. When users chat, Assistants decide which plugins to use to search for information, analyze data, & take action.
2. **Ambient Plugins (aka. Ambient Agents)** run in the background. System events & scheduled jobs "trigger" them, but they can also chat with users by sending notifications.

\


# Plugins

You build **Plugins** in Tool Studio.

## What type of plugin do I need?

The first architectural decision is simple: **What activates your plugin?**

```mermaid
flowchart TD
		B{Trigger?}
    B -->|System Event| C[Ambient Agent]
    B -->|User Utterance| D[Conversational Plugin]
    style C fill:#fff3e0
    style D fill:#e8f5e9
```

### User Utterance → Conversational Plugins

**Examples:**

* "Check my PTO balance"
* "Update the account owner in Salesforce"
* "Schedule a meeting with my team next week"

**Core components:**

* [**Natural Language Trigger**](/tool-studio/core-concepts/conversational-plugins/natural-language-triggers): Auto-generated from plugin name/description
* [**Conversation Process**](/tool-studio/conversation-process): Slots (aka. inputs) required from the user & a set of actions the Assistant should take

**Next steps:**

* Read best practices for [designing Conversational Plugins](/tool-studio/core-concepts/conversational-plugins).

\


### System Event → Ambient Agents

**When to use:** Your plugin responds to external system events (webhooks) or runs on a schedule (cron, interval, calendar).

**Examples:**

* Daily Customer Feedback Audit (scheduled)
* Expense Policy Violation Detection (webhook)
* Time Off Request Approvals (webhook + user notification)

**Core components:**

* **System Trigger**: Webhook listener or schedule
* **Compound Action**: Chains API calls without user interaction
* **Conversation Process** (optional): If you need to notify or interact with users

**Next steps:**

* Read best practices for [designing Ambient Agents](/ai-assistant/ambient-agents).