|
| 1 | +--- |
| 2 | +layout: page_v2 |
| 3 | +title: Salesagent |
| 4 | +description: A media sales agent that implements the AdCP Media Buy protocol |
| 5 | +sidebarType: 10 |
| 6 | +--- |
| 7 | + |
| 8 | +# Prebid Sales Agent |
| 9 | + |
| 10 | +The Prebid Sales Agent is a server that exposes advertising inventory to AI agents via the Model Context Protocol (MCP) and Agent-to-Agent (A2A) protocol. It is designed to integrate with ad servers like Google Ad Manager and provides tools for managing inventory and campaigns throughout their lifecycle. |
| 11 | + |
| 12 | +<div class="alert alert-info" role="alert"> |
| 13 | + For the full source code and latest updates, visit the <a href="https://github.com/prebid/salesagent">prebid/salesagent repository</a>. |
| 14 | +</div> |
| 15 | + |
| 16 | +## Key Features |
| 17 | + |
| 18 | +### For AI Agents |
| 19 | + |
| 20 | +- **Product Discovery**: Natural language search for advertising products. |
| 21 | +- **Campaign Creation**: Automated media buying with targeting capabilities. |
| 22 | +- **Creative Management**: Streamlined upload and approval workflows. |
| 23 | +- **Performance Monitoring**: Real-time access to campaign metrics. |
| 24 | + |
| 25 | +### For Publishers |
| 26 | + |
| 27 | +- **Multi-Tenant System**: Isolates data per publisher for security and organization. |
| 28 | +- **Adapter Pattern**: Supports multiple ad servers (e.g., Google Ad Manager). |
| 29 | +- **Real-time Dashboard**: Live activity feed powered by Server-Sent Events (SSE). |
| 30 | +- **Workflow Management**: Unified system for human-in-the-loop approvals. |
| 31 | +- **Admin Interface**: Web UI with Google OAuth for easy management. |
| 32 | + |
| 33 | +### For Developers |
| 34 | + |
| 35 | +- **MCP Protocol**: Standard interface for AI agents. |
| 36 | +- **A2A Protocol**: Agent-to-Agent communication via JSON-RPC 2.0. |
| 37 | +- **REST API**: Programmatic tenant management. |
| 38 | +- **Docker Deployment**: Easy setup for both local and production environments. |
| 39 | + |
| 40 | +## Getting Started |
| 41 | + |
| 42 | +### Quick Start (Evaluation) |
| 43 | + |
| 44 | +You can try the sales agent locally using Docker: |
| 45 | + |
| 46 | +```bash |
| 47 | +# Clone and start |
| 48 | +git clone https://github.com/prebid/salesagent.git |
| 49 | +cd salesagent |
| 50 | +docker compose up -d |
| 51 | + |
| 52 | +# Test the MCP interface |
| 53 | +uvx adcp http://localhost:8000/mcp/ --auth test-token list_tools |
| 54 | +``` |
| 55 | + |
| 56 | +Access services at [http://localhost:8000](http://localhost:8000): |
| 57 | + |
| 58 | +- **Admin UI**: `/admin` (Test credentials: `test123`) |
| 59 | +- **MCP Server**: `/mcp/` |
| 60 | +- **A2A Server**: `/a2a` |
| 61 | + |
| 62 | +### Production Deployment |
| 63 | + |
| 64 | +For production, publishers can deploy their own sales agent instance. The repository provides guides for various deployment methods, including Docker and cloud platforms. |
| 65 | + |
| 66 | +## The AdContext Protocol (AdCP) |
| 67 | + |
| 68 | +The Sales Agent is built on the **AdContext Protocol (AdCP)**, an open standard designed to standardize how AI agents interact with advertising platforms. |
| 69 | + |
| 70 | +<div class="alert alert-info" role="alert"> |
| 71 | + For comprehensive documentation, visit <a href="https://docs.adcontextprotocol.org/docs/intro">docs.adcontextprotocol.org</a>. |
| 72 | +</div> |
| 73 | + |
| 74 | +### Protocol Architecture |
| 75 | + |
| 76 | +AdCP operates as a layer on top of standard AI interaction protocols: |
| 77 | + |
| 78 | +- **MCP (Model Context Protocol)**: Facilitates direct integration with AI assistants (e.g., Claude Desktop). |
| 79 | +- **A2A (Agent-to-Agent Protocol)**: Enables complex, autonomous workflows and collaboration between agents using JSON-RPC 2.0. |
| 80 | + |
| 81 | +### Core Concepts |
| 82 | + |
| 83 | +AdCP abstracts complex advertising operations into standardized domains: |
| 84 | + |
| 85 | +1. **Inventory Discovery** (`get_products`): Agents can search for ad products using natural language criteria (e.g., "video ads in North America") rather than specific line item IDs. |
| 86 | +2. **Media Buying** (`create_media_buy`): A normalized workflow for proposal, negotiation, and booking that works consistently across different ad servers. |
| 87 | +3. **Creative Management** (`build_creative`): Standardized handling of creative assets, allowing agents to generate or upload assets that match publisher specifications. |
| 88 | +4. **Signal Activation** (`get_signals`, `activate_signal`): Mechanisms for passing context and identity signals to improve targeting and campaign performance. |
| 89 | + |
| 90 | +### Workflow Example |
| 91 | + |
| 92 | +A typical AI-driven campaign flow using AdCP might look like this: |
| 93 | + |
| 94 | +1. **Discovery**: Expected outcome is a list of available "Products" matching the agent's intent. |
| 95 | +2. **Planning**: The agent uses `create_media_buy` to submit a proposal. |
| 96 | +3. **Review**: The Sales Agent (and potentially a human publisher) reviews the proposal. |
| 97 | +4. **Execution**: Once approved, the Sales Agent pushes the orders to the underlying ad server (e.g., GAM). |
| 98 | + |
| 99 | +## Architecture |
| 100 | + |
| 101 | +The project follows a clean structure isolating core MCP components, business logic services, and ad server adapters. |
| 102 | + |
| 103 | +```text |
| 104 | +salesagent/ |
| 105 | +├── src/ |
| 106 | +│ ├── core/ # Core MCP server components |
| 107 | +│ ├── services/ # Business logic services |
| 108 | +│ ├── adapters/ # Ad server integrations (e.g., GAM) |
| 109 | +│ └── admin/ # Admin UI (Flask) |
| 110 | +├── scripts/ # Utility and deployment scripts |
| 111 | +└── tests/ # Comprehensive test suite |
| 112 | +``` |
| 113 | + |
| 114 | +## Contributing |
| 115 | + |
| 116 | +Contributions are welcome! Please refer to the [Development Guide](https://github.com/prebid/salesagent/blob/main/docs/development/README.md) in the repository for details on setting up your environment and creating pull requests. |
0 commit comments