> ## Documentation Index
> Fetch the complete documentation index at: https://agenticadvertisingorg-snap-format-preview-links.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Platform / intermediary track

> AdCP platform track (D1-D4): MCP server architecture, supply path verification, agent trust, RTB-to-AdCP migration patterns, and an infrastructure build project.

# Platform / intermediary track (D1–D4)

<Info>
  **Members only** — Requires Basics credential (A1–A3). Four modules, \~105 minutes total.
</Info>

This track is for people building AdCP infrastructure: ad tech platforms, exchanges, data companies, and anyone connecting the ecosystem. You'll learn MCP server architecture, supply chain verification, RTB migration, and build working infrastructure in the build project.

Completing this track (plus A1–A3) earns the **AdCP practitioner** credential.

***

## D1: MCP server architecture

**\~20 min** | Prerequisite: A3

Technical deep dive on building an AdCP-compliant MCP server. Transport options (SSE, Streamable HTTP), tool definition patterns, OAuth/authorization flows, and account management.

### Reading list

<CardGroup cols={2}>
  <Card title="Building with AdCP" icon="wrench" href="/docs/building/index">
    Overview of what it takes to build an AdCP integration.
  </Card>

  <Card title="MCP integration guide" icon="plug" href="/docs/building/integration/mcp-guide">
    The definitive guide: tool calls, response format, available tools, context management.
  </Card>

  <Card title="Authentication" icon="lock" href="/docs/building/integration/authentication">
    OAuth 2.0 for agent authentication, token management, and operator credentials.
  </Card>

  <Card title="Accounts and agents" icon="people-arrows" href="/docs/building/integration/accounts-and-agents">
    How brands, operators, agents, and accounts relate to each other.
  </Card>

  <Card title="Context sessions" icon="timeline" href="/docs/building/integration/context-sessions">
    Managing context across multi-turn agent interactions.
  </Card>

  <Card title="Schemas and SDKs" icon="box" href="/docs/building/schemas-and-sdks">
    JSON Schemas, TypeScript types, and client SDKs for building faster.
  </Card>
</CardGroup>

### Key concepts

* **MCP server architecture** — exposing AdCP tasks as tools, handling auth, request routing
* **Transport options** — Streamable HTTP for most cases, SSE for real-time updates
* **Capability advertising** — `get_adcp_capabilities` so other agents know what you support
* **Account handling** — managing incoming `sync_accounts` from buyers

<Card title="Start D1 with Addie" icon="play" href="https://agenticadvertising.org/chat">
  "I'd like to start certification module D1."
</Card>

***

## D2: Supply path, trust, and property governance

**\~20 min** | Prerequisite: D1

Cryptographic signatures for supply chain verification. How platforms validate agent identity, detect fraud, and ensure trust. The relationship between AdCP and ads.cert.

### Reading list

<CardGroup cols={2}>
  <Card title="Property governance" icon="shield-check" href="/docs/governance/property/index">
    Identity, authorization, and data enrichment for participants in AdCP.
  </Card>

  <Card title="AdAgents.json" icon="file-code" href="/docs/governance/property/adagents">
    Publisher authorization for which agents may sell or enrich inventory.
  </Card>

  <Card title="Seller setup" icon="store" href="/docs/brand-protocol/seller-setup">
    How brand.json and adagents.json work together for seller identity and supply path verification.
  </Card>

  <Card title="Seller verification" icon="shield-halved" href="/docs/verification/overview">
    Buyer-side walkthrough of signatures, seller brand.json, publisher adagents.json, and mutual assertion.
  </Card>

  <Card title="Authorized properties" icon="key" href="/docs/governance/property/authorized-properties">
    How publishers declare authorized sellers, and how buyers verify them.
  </Card>

  <Card title="Security" icon="shield" href="/docs/building/implementation/security">
    Security implementation patterns for AdCP servers.
  </Card>

  <Card title="Accounts and security" icon="user-shield" href="/docs/media-buy/advanced-topics/accounts-and-security">
    Account-level security, agent and account hierarchies, and access control.
  </Card>

  <Card title="Property specification" icon="scroll" href="/docs/governance/property/specification">
    The formal specification for property governance.
  </Card>

  <Card title="Campaign governance" icon="shield-check" href="/docs/governance/campaign">
    Multi-party validation: how platforms implement the governance flow for autonomous transactions.
  </Card>

  <Card title="Campaign governance specification" icon="file-code" href="/docs/governance/campaign/specification">
    The full technical specification: sync\_plans, check\_governance, report\_plan\_outcome, audit logs.
  </Card>

  <Card title="Policy registry" icon="book" href="/docs/governance/policy-registry">
    Community-maintained compliance policies that governance agents resolve and enforce.
  </Card>
</CardGroup>

### Key concepts

* **Agent identity verification** — seller `brand.json` declares the operator, sales endpoints, and signing-key discovery; publisher `adagents.json` declares which agents are authorized for which properties
* **Supply path transparency** — buyer agents verify both sides of the relationship: the operator claims the sales path in `brand.json`, and the publisher confirms it in `adagents.json`
* **ads.cert relationship** — extending cryptographic verification from RTB to agent-to-agent interactions
* **Campaign governance architecture** — implement three-party validation: sync governance agents via `sync_governance`, call `check_governance` before executing, handle all statuses
* **Policy registry integration** — resolve policies by ID, integrate natural language policy text and exemplars into governance agent evaluation

<Card title="Start D2 with Addie" icon="play" href="https://agenticadvertising.org/chat">
  "I'd like to start certification module D2."
</Card>

***

## D3: RTB coexistence and migration

**\~20 min** | Prerequisite: D2

How AdCP coexists with existing programmatic infrastructure. Migration strategies for DSPs, SSPs, and exchanges. Running parallel systems during transition.

### Reading list

<CardGroup cols={2}>
  <Card title="Trusted Match Protocol" icon="microchip" href="/docs/trusted-match">
    The real-time execution layer: context match, identity match, and cross-publisher activation.
  </Card>

  <Card title="AdCP and OpenRTB" icon="arrows-left-right" href="/docs/building/understanding/adcp-vs-openrtb">
    The clearest view of how AdCP campaign workflows connect to impression-time execution.
  </Card>

  <Card title="Async operations" icon="hourglass-half" href="/docs/building/implementation/async-operations">
    Handling long-running operations — essential for bridging real-time and agentic systems.
  </Card>

  <Card title="Webhooks" icon="bell" href="/docs/building/implementation/webhooks">
    Event-driven updates for delivery, status changes, and campaign modifications.
  </Card>

  <Card title="Error handling" icon="triangle-exclamation" href="/docs/building/implementation/error-handling">
    Error patterns, retry strategies, and graceful degradation.
  </Card>

  <Card title="Sandbox" icon="flask" href="/docs/media-buy/advanced-topics/sandbox">
    Testing your implementation against sandbox agents before going live.
  </Card>

  <Card title="Implementor FAQ" icon="circle-question" href="/docs/reference/implementor-faq">
    Common questions and answers from teams building AdCP integrations.
  </Card>

  <Card title="Router architecture" icon="diagram-project" href="/docs/trusted-match/router-architecture">
    Deployment, fan-out, and provider configuration for the TMP Router.
  </Card>

  <Card title="AI mediation" icon="microchip-ai" href="/docs/trusted-match/ai-mediation">
    How demand reaches AI assistants through the Trusted Match Protocol.
  </Card>
</CardGroup>

### Key concepts

* **Coexistence strategy** — run AdCP alongside OpenRTB, gradually migrating workflows
* **TMP as the bridge** — TMP connects planning-time media buys to serve-time decisions through two structurally separated operations (Context Match and Identity Match). The TMP Router fans out to configured providers in parallel and the publisher joins responses locally. This is how cross-publisher frequency capping, suppression, and suitability work at impression time. TMP is an [experimental surface](/docs/reference/experimental-status) in AdCP 3.0 (feature id `trusted_match.core`); router, exposure-token, and identity-match details may change between 3.x releases with at least 6 weeks' notice.
* **Router deployment and fan-out patterns** — how the TMP Router is deployed, provider configuration, and parallel fan-out to multiple match providers
* **Platform-specific migration** — DSPs wrap bidding logic, SSPs expose inventory, exchanges translate
* **Performance benchmarking** — compare agentic vs traditional on campaign performance, efficiency, cost

<Card title="Start D3 with Addie" icon="play" href="https://agenticadvertising.org/chat">
  "I'd like to start certification module D3."
</Card>

***

## D4: Build project — AdCP infrastructure

**\~45 min** | Prerequisite: D3

Build working AdCP infrastructure using any AI coding assistant (Claude Code, Cursor, Copilot) with the [`@adcp/sdk`](/docs/building/schemas-and-sdks) SDK. Point your coding assistant at a [skill file](/docs/building/build-an-agent) — choose `build-seller-agent` or `build-signals-agent` depending on what you want to build. This is the most ambitious build project.

### What you'll build

* MCP server with AdCP tool definitions
* `get_adcp_capabilities` implementation
* At least 3 AdCP tasks exposed as tools
* OAuth/authentication flow
* Proper error handling with AdCP error codes

### How you'll validate

Run the matching storyboard against your running agent:

```bash theme={null}
npx @adcp/sdk@latest storyboard run my-agent media_buy_seller
```

The storyboard validates protocol compliance across the complete workflow. See [Validate Your Agent](/docs/building/validate-your-agent) for setup, debugging, and the full CLI reference.

### Specialisms you can claim

Platform agents often span multiple domains. Declare each protocol you implement in `supported_protocols` and each specialism you claim in `specialisms`. Some common combinations:

* Full-stack sales platform: `supported_protocols: ["media_buy", "creative", "signals"]`, `specialisms: ["sales-guaranteed", "sales-non-guaranteed", "creative-ad-server"]`
* Signal platform: `supported_protocols: ["signals"]`, `specialisms: ["signal-marketplace"]` or `["signal-owned"]`
* Governance platform: `supported_protocols: ["governance"]`, `specialisms: ["content-standards", "property-lists", "collection-lists"]`

See the [Compliance Catalog](/docs/building/compliance-catalog) for every specialism and the storyboards that verify each claim.

### Assessment rubric

| Dimension             | Weight | What Addie evaluates                                |
| --------------------- | ------ | --------------------------------------------------- |
| Specification quality | 20%    | Can you specify infrastructure in AdCP terms?       |
| Schema compliance     | 20%    | Protocol compliance across all endpoints            |
| Error handling        | 15%    | Proper error handling with recovery types and async |
| Design rationale      | 25%    | Can you reason about production architecture?       |
| Extension ability     | 20%    | Can you extend the endpoint with new tasks?         |

Passing threshold: 70%.

<Card title="Start D4 with Addie" icon="play" href="https://agenticadvertising.org/chat">
  "I'd like to start certification module D4."
</Card>

***

## What's next

After completing D1–D4, you've earned the **AdCP practitioner** credential. From here you can pursue specialist modules:

* [S1: Media buy](/docs/learning/specialist/media-buy) — transaction flows, pricing, orchestration
* [S2: Creative](/docs/learning/specialist/creative) — asset workflows, format compliance
* [S3: Signals](/docs/learning/specialist/signals) — measurement, attribution, optimization
* [S4: Governance](/docs/learning/specialist/governance) — brand safety, campaign governance, compliance, policy registry
* [S5: Sponsored Intelligence](/docs/learning/specialist/sponsored-intelligence) — conversational brand experiences