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

# Introduction to Agents

> Build, version, and operationalize AI agents in Bud AI Foundry

## Overview

Bud AI Foundry Agents help teams design reusable prompt-driven assistants with deployment-aware configuration, tool connectivity, and version lifecycle controls.

```mermaid theme={null}
graph LR
    A[Design Agent] --> B[Configure Model or Deployment]
    B --> C[Attach Tools and Schemas]
    C --> D[Test and Save]
    D --> E[Version and Reuse]
```

Agents can be managed from the global **Agents** catalog or from each project's **Agents** tab, so platform teams and project owners can collaborate from the same source of truth.

## Key Capabilities

**Unified Discovery**
Search and filter agents by name, type, category, author, and tags.

**Guided Creation Workflow**
Create agents through a multi-step drawer with project context, type selection, model/deployment selection, and prompt configuration.

**Structured + Unstructured Inputs**
Support schema-backed input/output variables for predictable integrations, or freeform prompts for exploratory usage.

**Built-in Versioning**
Use the Versions tab to inspect current and previous versions, add versions, and set defaults.

**Tool Integrations (MCP Connectors)**
Attach connectors and tools directly inside the agent workflow for end-to-end orchestration.

**Native Tools**
Enable first-party tools — Web Search, Web Fetch, and Code Interpreter — on a prompt version with declarative configuration. No credentials or connector setup required.

## Where Agents Fit

| Surface                  | Primary Use                                                           |
| ------------------------ | --------------------------------------------------------------------- |
| **Global Agents Page**   | Catalog browsing, filtering, and cross-project discovery              |
| **Project Agents Tab**   | Project-scoped creation, sorting, and lifecycle actions               |
| **Agent Details**        | Overview, integration snippets, and version management                |
| **Agent Builder Drawer** | Multi-step creation/editing, schema configuration, tools, and testing |

## Typical Lifecycle

1. Create a workflow from the Agents page or project Agents tab.
2. Configure prompt behavior, deployment target, and optional schemas.
3. Connect tools if external actions are needed.
4. Run and validate behavior.
5. Save and publish.
6. Add and manage versions over time.

## Next Steps

<CardGroup cols={3}>
  <Card title="Quick Start" icon="play" href="/prompts-agents/quickstart">
    Create your first agent in minutes
  </Card>

  <Card title="Agent Concepts" icon="book" href="/prompts-agents/agent-concepts">
    Understand sessions, versions, and tools
  </Card>

  <Card title="Build Tutorial" icon="graduation-cap" href="/prompts-agents/creating-first-agent">
    Follow a step-by-step creation flow
  </Card>
</CardGroup>