Build Seller Agent

---

name: build-seller-agent description: Use when building an AdCP seller agent — a publisher, SSP, or retail media network that sells advertising inventory to buyer agents.

Build a Seller Agent

Overview

A seller agent receives briefs from buyers, returns products with pricing, accepts media buys, manages creatives, and reports delivery. The business model — what you sell, how you price it, and whether humans approve deals — shapes every implementation decision. Determine that first.

When to Use

• User wants to build an agent that sells ad inventory
• User mentions publisher, SSP, retail media, or media network in the context of AdCP
• User references get_products, create_media_buy, or the media buy protocol

Not this skill:

• Buying ad inventory → that's a buyer/DSP agent (see docs/getting-started.md)
• Serving audience segments → skills/build-signals-agent/
• Rendering creatives from briefs → that's a creative agent

Before Writing Code

Determine these five things. Ask the user — don't guess.

1. What Kind of Seller?

• Premium publisher — guaranteed inventory, fixed pricing, IO approval (ESPN, NYT)
• SSP / Exchange — non-guaranteed, auction-based, instant activation
• Retail media network — both guaranteed and non-guaranteed, proposals, catalog-driven creative, conversion tracking

2. Guaranteed or Non-Guaranteed?

• Guaranteed — delivery_type: "guaranteed", may require async approval (submitted → pending_approval → confirmed)
• Non-guaranteed — delivery_type: "non_guaranteed", buyer sets bid_price, instant activation

Many sellers support both — different products can have different delivery types.

3. Products and Pricing

Get specific inventory. Each product needs:

• product_id, name, description
• publisher_properties — at least one { publisher_domain: 'example.com', selection_type: 'all' } (discriminated union: 'all' | 'by_id' with property_ids | 'by_tag' with tags)
• format_ids — array of { agent_url: string, id: string } referencing creative formats
• delivery_type — 'guaranteed' or 'non_guaranteed'
• pricing_options — at least one (see below)
• reporting_capabilities — { available_reporting_frequencies: ['daily'], expected_delay_minutes: 240, timezone: 'UTC', supports_webhooks: false, available_metrics: ['impressions', 'spend', 'clicks'], date_range_support: 'date_range' }
• Optional: channels — use as const to avoid string[] inference: channels: ['display', 'olv'] as const

Pricing models (all require pricing_option_id and currency):

• cpm — { pricing_option_id: 'cpm-1', pricing_model: "cpm", fixed_price: 12.00, currency: "USD" }
• cpc — { pricing_option_id: 'cpc-1', pricing_model: "cpc", fixed_price: 1.50, currency: "USD" }
• Auction — { pricing_option_id: 'auction-1', pricing_model: "cpm", floor_price: 5.00, currency: "USD" } (buyer bids above floor)

Each pricing option can set min_spend_per_package to enforce minimum budgets.

For all PricingOption variants and Product required fields, see docs/TYPE-SUMMARY.md.

4. Approval Workflow

For guaranteed buys, choose one:

• Instant confirmation — create_media_buy returns completed with confirmed status. Simplest.
• Async approval — returns submitted, buyer polls get_media_buys. Use registerAdcpTaskTool.
• Human-in-the-loop — returns input-required with a setup URL for IO signing.

Non-guaranteed buys are always instant confirmation.

5. Creative Management

• Standard — list_creative_formats + sync_creatives. Buyer uploads assets, seller validates.
• Catalog-driven — buyer syncs product catalog via sync_catalogs. Common for retail media.
• None — creative handled out-of-band. Omit creative tools.

Tools and Required Response Shapes

get_adcp_capabilities — register first, empty {} schema

capabilitiesResponse({
  adcp: { major_versions: [3] },
  supported_protocols: ['media_buy'],
})

sync_accounts — SyncAccountsRequestSchema.shape

taskToolResponse({
  accounts: [{
    account_id: string,       // required - your platform's ID
    brand: { domain: string },// required - echo back from request
    operator: string,         // required - echo back from request
    action: 'created' | 'updated',  // required
    status: 'active' | 'pending_approval',  // required
  }]
})

sync_governance — SyncGovernanceRequestSchema.shape

taskToolResponse({
  accounts: [{
    account: { brand: {...}, operator: string },  // required - echo back
    status: 'synced',         // required
    governance_agents: [{ url: string, categories?: string[] }],  // required
  }]
})

get_products — GetProductsRequestSchema.shape

productsResponse({
  products: [{
    product_id: 'prod-1',
    name: 'Homepage Display',
    description: 'Premium display ads on homepage',
    publisher_properties: [{ publisher_domain: 'example.com', selection_type: 'all' }],
    format_ids: [{ agent_url: 'https://creative.example.com/mcp', id: 'display-300x250' }],
    delivery_type: 'guaranteed',
    pricing_options: [{
      pricing_option_id: 'cpm-standard',
      pricing_model: 'cpm',
      fixed_price: 12.00,
      currency: 'USD',
    }],
    reporting_capabilities: {
      available_reporting_frequencies: ['daily'],
      expected_delay_minutes: 240,
      timezone: 'UTC',
      supports_webhooks: false,
      available_metrics: ['impressions', 'spend', 'clicks'],
      date_range_support: 'date_range',
    },
  }],
  sandbox: true,        // for mock data
})

create_media_buy — CreateMediaBuyRequestSchema.shape

Validate the request before creating the buy. Return an error response (not adcpError) when business validation fails:

// Success — revision, confirmed_at, and valid_actions are auto-set:
mediaBuyResponse({
  media_buy_id: string,       // required
  status: 'pending_creatives',// triggers valid_actions auto-population
  packages: [{                // required
    package_id: string,
    product_id: string,
    pricing_option_id: string,
    budget: number,
  }],
})

// Validation failure (reversed dates, budget too low, unknown product):
adcpError('INVALID_REQUEST', { message: 'start_time must be before end_time' })

get_media_buys — GetMediaBuysRequestSchema.shape

getMediaBuysResponse({
  media_buys: [{
    media_buy_id: string,   // required
    status: 'active' | 'pending_start' | ...,  // required
    currency: 'USD',        // required
    confirmed_at: string,   // required for guaranteed approval — ISO timestamp
    packages: [{
      package_id: string,   // required
    }],
  }]
})

list_creative_formats — ListCreativeFormatsRequestSchema.shape

listCreativeFormatsResponse({
  formats: [{
    format_id: { agent_url: string, id: string },  // required
    name: string,  // required
  }]
})

sync_creatives — SyncCreativesRequestSchema.shape

syncCreativesResponse({
  creatives: [{
    creative_id: string,          // required - echo from request
    action: 'created' | 'updated',  // required
  }]
})

get_media_buy_delivery — GetMediaBuyDeliveryRequestSchema.shape

deliveryResponse({
  reporting_period: { start: string, end: string },  // required - ISO timestamps
  media_buy_deliveries: [{
    media_buy_id: string,     // required
    status: 'active',         // required
    totals: { impressions: number, spend: number },  // required
    by_package: [],           // required (can be empty)
  }]
})

Context and Ext Passthrough

Every AdCP request includes an optional context field. Buyers use it to carry correlation IDs, orchestration metadata, and workflow state across multi-agent calls. Your agent must echo the context object back unchanged in every response.

// In every tool handler:
const context = args.context; // may be undefined — that's fine

// In every response:
return taskToolResponse({
  // ... your response fields ...
  context, // echo it back unchanged
});

Do not modify, inspect, or omit the context — treat it as opaque. If the request has no context, omit it from the response.

Some schemas also define an ext field for vendor-namespaced extensions. If your request schema includes ext, accept it without error. Tools with explicit ext support: sync_governance, provide_performance_feedback, sync_event_sources.

Compliance Testing (Optional)

Add registerTestController so the comply framework can deterministically test your state machines. Without it, compliance testing relies on observational storyboards that can't force state transitions.

import { registerTestController } from '@adcp/sdk';
import type { TestControllerStore } from '@adcp/sdk';

const store: TestControllerStore = {
  async forceAccountStatus(accountId, status) {
    const prev = accounts.get(accountId);
    if (!prev) throw new TestControllerError('NOT_FOUND', `Account ${accountId} not found`);
    accounts.set(accountId, status);
    return { success: true, previous_state: prev, current_state: status };
  },
  async forceMediaBuyStatus(mediaBuyId, status) {
    const prev = mediaBuys.get(mediaBuyId);
    if (!prev) throw new TestControllerError('NOT_FOUND', `Media buy ${mediaBuyId} not found`);
    const terminal = ['completed', 'rejected', 'canceled'];
    if (terminal.includes(prev))
      throw new TestControllerError('INVALID_TRANSITION', `Cannot transition from ${prev}`, prev);
    mediaBuys.set(mediaBuyId, status);
    return { success: true, previous_state: prev, current_state: status };
  },
  async forceCreativeStatus(creativeId, status, rejectionReason) {
    const prev = creatives.get(creativeId);
    if (!prev) throw new TestControllerError('NOT_FOUND', `Creative ${creativeId} not found`);
    // archived blocks transitions to active states, but archived → rejected is valid (compliance override)
    const activeStatuses = ['processing', 'pending_review', 'approved'];
    if (prev === 'archived' && activeStatuses.includes(status))
      throw new TestControllerError('INVALID_TRANSITION', `Cannot transition from archived to ${status}`, prev);
    creatives.set(creativeId, status);
    return { success: true, previous_state: prev, current_state: status };
  },
  async simulateDelivery(mediaBuyId, params) {
    // params: { impressions?: number, clicks?: number, reported_spend?: { amount, currency }, conversions?: number }
    return { success: true, simulated: { ...params }, cumulative: { ...params } };
  },
  async simulateBudgetSpend(params) {
    return { success: true, simulated: { spend_percentage: params.spend_percentage } };
  },
};

registerTestController(server, store);

When using this, declare compliance_testing in supported_protocols:

capabilitiesResponse({
  adcp: { major_versions: [3] },
  supported_protocols: ['media_buy', 'compliance_testing'],
})

Only implement the store methods for scenarios your agent supports. Unimplemented methods are excluded from list_scenarios automatically.

The storyboard tests state machine correctness:

• NOT_FOUND when forcing transitions on unknown entities
• INVALID_TRANSITION when transitioning from terminal states (completed, rejected, canceled for media buys; archived blocks active states like processing/pending_review/approved, but archived → rejected is valid)
• Successful transitions between valid states

Throw TestControllerError from store methods for typed errors. The SDK validates status enum values before calling your store.

Validate with: adcp storyboard run <agent> deterministic_testing --json

Session-backed stores (factory shape)

Don't close over module-scoped maps. If your session state is persisted (Postgres, Redis, JSONB) and rehydrated into a new object per request, a store whose methods close over a module-level WeakMap<SessionState, …> or module-scoped cache will silently drop entries between calls — the cached ref was GC'd when the session was serialized out and rebuilt.

Use the factory shape. scenarios declares the static capability set — the SDK answers list_scenarios from this field and never invokes createStore for capability probes, so it's safe to throw on missing session_id. createStore runs per request for every other scenario, returning a store bound to the live session.

import {
  registerTestController,
  CONTROLLER_SCENARIOS,
  enforceMapCap,
  TestControllerError,
} from '@adcp/sdk';

registerTestController(server, {
  scenarios: [
    CONTROLLER_SCENARIOS.FORCE_ACCOUNT_STATUS,
    CONTROLLER_SCENARIOS.FORCE_MEDIA_BUY_STATUS,
    CONTROLLER_SCENARIOS.FORCE_CREATIVE_STATUS,
    CONTROLLER_SCENARIOS.SIMULATE_DELIVERY,
    CONTROLLER_SCENARIOS.SIMULATE_BUDGET_SPEND,
  ],
  async createStore(input) {
    const sessionId = (input.context as { session_id?: string })?.session_id;
    if (!sessionId) throw new TestControllerError('INVALID_PARAMS', 'context.session_id is required');
    const session = await loadSession(sessionId);

    return {
      async forceAccountStatus(accountId, status) {
        // enforceMapCap only rejects NET-NEW keys at the cap; updating an
        // existing accountId always passes, so calling it before every set()
        // is safe.
        enforceMapCap(session.accountStatuses, accountId, 'account statuses');
        const prev = session.accountStatuses.get(accountId) ?? 'active';
        session.accountStatuses.set(accountId, status);
        await saveSession(session);
        return { success: true, previous_state: prev, current_state: status };
      },

      async forceMediaBuyStatus(mediaBuyId, status) {
        const prev = session.mediaBuyStatuses.get(mediaBuyId);
        if (!prev) throw new TestControllerError('NOT_FOUND', `Media buy ${mediaBuyId} not found`);
        const terminal = ['completed', 'rejected', 'canceled'];
        if (terminal.includes(prev)) {
          throw new TestControllerError('INVALID_TRANSITION', `Cannot transition from ${prev}`, prev);
        }
        enforceMapCap(session.mediaBuyStatuses, mediaBuyId, 'media buy states');
        session.mediaBuyStatuses.set(mediaBuyId, status);
        await saveSession(session);
        return { success: true, previous_state: prev, current_state: status };
      },

      // ...implement other scenarios from your `scenarios` list the same way
    };
  },
});

Cap per-session maps

Wrap every Map.set on session-scoped state with enforceMapCap to reject unbounded growth with a typed INVALID_STATE error (vs. silent LRU eviction, which would make compliance tests nondeterministic). Existing-key overwrites always pass — only net-new keys are rejected at the cap. Default cap is SESSION_ENTRY_CAP (1000).

Custom MCP wrappers

If you need AsyncLocalStorage, sandbox gating, or a custom task store around the controller tool, bypass registerTestController and call the exported building blocks directly. toMcpResponse and TOOL_INPUT_SHAPE are the exact pieces the default registration uses — reusing them keeps the envelope shape identical.

import { AsyncLocalStorage } from 'node:async_hooks';
import { handleTestControllerRequest, toMcpResponse, TOOL_INPUT_SHAPE } from '@adcp/sdk';

const sessionContext = new AsyncLocalStorage<{ sessionId: string }>();
const store = { async forceAccountStatus(id, status) { /* ... */ } };

server.tool('comply_test_controller', 'Sandbox only.', TOOL_INPUT_SHAPE, async input => {
  if (!sandboxEnabled()) {
    return toMcpResponse({ success: false, error: 'FORBIDDEN', error_detail: 'Sandbox disabled' });
  }
  const sessionId = (input.context as { session_id: string }).session_id;
  return sessionContext.run({ sessionId }, async () => {
    const response = await handleTestControllerRequest(store, input as Record<string, unknown>);
    return toMcpResponse(response);
  });
});

SDK Quick Reference

SDK piece	Usage
createAdcpServerFromPlatform(platform, opts)	Build a server from a typed DecisioningPlatform — compile-time specialism enforcement, ctx_metadata round-trip, idempotency-principal synthesis, status mappers, webhook auto-emit
createAdcpServer(config) (legacy)	v5 handler-bag entry. Mid-migration / escape-hatch only; reach via @adcp/sdk/server/legacy/v5
serve(() => createAdcpServerFromPlatform(platform, opts))	Start HTTP server on :3001/mcp
ctx.store	State store in every handler — get, put, patch, delete, list
InMemoryStateStore	Default state store (dev/testing)
PostgresStateStore	Production state store (shared across instances)
DEFAULT_REPORTING_CAPABILITIES	Use as reporting_capabilities: DEFAULT_REPORTING_CAPABILITIES on products
checkGovernance(options)	Call governance agent before financial commits
governanceDeniedError(result)	Convert governance denial to GOVERNANCE_DENIED error
mediaBuyResponse(data)	Auto-applied for createMediaBuy (sets revision, confirmed_at, valid_actions)
adcpError(code, { message })	Structured error (e.g., BUDGET_TOO_LOW, PRODUCT_NOT_FOUND)
registerTestController(server, store | { scenarios, createStore })	Add comply_test_controller. Plain store or per-request factory.
TestControllerError(code, message)	Typed error from store methods
handleTestControllerRequest(store, input)	Low-level dispatch for custom MCP wrappers
toMcpResponse(response) / TOOL_INPUT_SHAPE	MCP envelope + Zod input schema for custom wrappers
enforceMapCap(map, key, label, cap?)	Reject net-new keys once a session Map hits SESSION_ENTRY_CAP (1000)
expectControllerError(result, code) / expectControllerSuccess(result)	Unit-test assertions — narrow responses to error or success arms

Response builders (productsResponse, mediaBuyResponse, deliveryResponse, etc.) are auto-applied by the framework — you return the data, the framework wraps it. You only need to call them directly for tools without a dedicated builder.

Import everything from @adcp/sdk/server. Types from @adcp/sdk/server with import type.

Setup

npm init -y
npm install @adcp/sdk
npm install -D typescript @types/node

Minimal tsconfig.json:

{
  "compilerOptions": {
    "target": "ES2022",
    "module": "Node16",
    "moduleResolution": "Node16",
    "strict": true,
    "skipLibCheck": true,
    "outDir": "dist"
  }
}

skipLibCheck: true avoids false-positive errors from transitive .d.ts files (e.g., @opentelemetry/api).

Implementation

Use createAdcpServerFromPlatform — it auto-wires schemas, response builders, and get_adcp_capabilities from a typed DecisioningPlatform class. Handlers receive (params, ctx) where ctx.store persists state, ctx.account is the resolved account, and ctx.ctxMetadata is the resource-keyed cache.

LEGACY (v5) — the worked example below is the v5 handler-bag shape (createAdcpServer({ accounts, mediaBuy })). It still compiles via @adcp/sdk/server/legacy/v5. For new agents, refer to skills/build-seller-agent/SKILL.md lines 44–85 (the canonical v6 example) or examples/decisioning-platform-programmatic.ts for the typed DecisioningPlatform class shape. Lift the handler bodies (governance check, randomUUID-minted ids, ctx.store use) into class MySeller implements DecisioningPlatform<{}, MyMeta> with accounts/sales fields, then createAdcpServerFromPlatform(new MySeller(), { name, version, idempotency }).

Imports: most things live at @adcp/sdk. The idempotency store helpers (createIdempotencyStore, memoryBackend, pgBackend) live at the narrower @adcp/sdk/server subpath. Both are re-exported from the root — either works — but splitting them makes intent obvious.

import { randomUUID } from 'node:crypto';
import {
  createAdcpServer,
  serve,
  adcpError,
  InMemoryStateStore,
  checkGovernance,
  governanceDeniedError,
} from '@adcp/sdk';
import { createIdempotencyStore, memoryBackend } from '@adcp/sdk/server';
import type { ServeContext } from '@adcp/sdk';

const stateStore = new InMemoryStateStore(); // shared across requests

// Idempotency — required for any v3-compliant seller that accepts mutating
// requests. `createIdempotencyStore` throws if `ttlSeconds` is outside the
// spec bounds (3600–604800).
const idempotency = createIdempotencyStore({
  backend: memoryBackend(), // pgBackend(pool) for production
  ttlSeconds: 86400, // 24 hours
});

function createAgent({ taskStore }: ServeContext) {
  return createAdcpServer({
    name: 'My Seller Agent',
    version: '1.0.0',
    taskStore,
    stateStore,
    idempotency,

    // Principal scoping for idempotency. MUST never return undefined — or
    // every mutating request rejects as SERVICE_UNAVAILABLE. A constant is
    // fine for a demo; for multi-tenant production use ctx.account typed
    // via the framework constructor's `<MyAccount>` generic.
    resolveSessionKey: () => 'default-principal',

    resolveAccount: async ref => {
      if ('account_id' in ref) return stateStore.get('accounts', ref.account_id);
      return null;
    },

    accounts: {
      syncAccounts: async (params, ctx) => {
        /* ... */
      },
    },
    mediaBuy: {
      getProducts: async (params, ctx) => {
        return { products: PRODUCTS, sandbox: true };
        // productsResponse() auto-applied by framework
      },
      createMediaBuy: async (params, ctx) => {
        // Governance check for financial commitment
        if (ctx.account?.governanceUrl) {
          const gov = await checkGovernance({
            agentUrl: ctx.account.governanceUrl,
            planId: params.plan_id ?? 'default',
            caller: 'https://my-agent.com/mcp',
            tool: 'create_media_buy',
            payload: params,
          });
          if (!gov.approved) return governanceDeniedError(gov);
        }
        // Use randomUUID (not Date.now) so ids are unguessable — a guessable
        // media_buy_id lets another buyer probe or cancel. Same applies to
        // any seller-issued id (package_id, creative_id, etc.).
        const buy = {
          media_buy_id: `mb_${randomUUID()}`,
          status: 'pending_creatives' as const,
          packages:
            params.packages?.map(pkg => ({
              package_id: `pkg_${randomUUID()}`,
              product_id: pkg.product_id,
              pricing_option_id: pkg.pricing_option_id,
              budget: pkg.budget,
            })) ?? [],
        };
        await ctx.store.put('media_buys', buy.media_buy_id, buy);
        return buy; // mediaBuyResponse() auto-applied (sets revision, confirmed_at, valid_actions)
      },
      updateMediaBuy: async (params, ctx) => {
        const existing = await ctx.store.get('media_buys', params.media_buy_id);
        if (!existing) {
          return adcpError('MEDIA_BUY_NOT_FOUND', {
            message: `No media buy with id ${params.media_buy_id}`,
            field: 'media_buy_id',
          });
        }
        // Only merge the fields you want to persist — do NOT spread `params`
        // wholesale. `params` carries envelope fields (idempotency_key,
        // context) that have no business in your domain state. Spreading
        // them pollutes `get_media_buys` responses and breaks dedup.
        const updated = { ...existing, status: params.active === false ? 'paused' : 'active' };
        await ctx.store.put('media_buys', params.media_buy_id, updated);
        return {
          media_buy_id: params.media_buy_id,
          status: updated.status as 'paused' | 'active',
          affected_packages: [],
        };
      },
      getMediaBuys: async (params, ctx) => {
        const result = await ctx.store.list('media_buys');
        return { media_buys: result.items };
      },
      getMediaBuyDelivery: async (params, ctx) => {
        /* ... */
      },
      listCreativeFormats: async (params, ctx) => {
        /* ... */
      },
      syncCreatives: async (params, ctx) => {
        return {
          // Response shape is `creatives: [{ creative_id, action }]` per the
          // sync_creatives response schema — NOT `synced_creatives`.
          creatives:
            params.creatives?.map(c => ({
              creative_id: c.creative_id ?? `cr_${randomUUID()}`,
              action: 'created' as const,
            })) ?? [],
        };
      },
    },
    capabilities: {
      features: { inlineCreativeManagement: false },
    },
  });
}

serve(createAgent);

Key points:

1. Single .ts file — one DecisioningPlatform class passed to createAdcpServerFromPlatform
2. get_adcp_capabilities is auto-generated from your handlers — don't register it manually (idempotency capability is auto-declared too)
3. Response builders are auto-applied — just return the data
4. Use ctx.store for state — persists across stateless HTTP requests
5. Set sandbox: true on all mock/demo responses
6. Use adcpError() for business validation failures
7. Use as const on string literal arrays and union-typed fields in product definitions — TypeScript infers string[] from ['display', 'olv'] but the SDK requires specific union types like MediaChannel[]. Apply as const to channels, delivery_type, selection_type, and pricing_model values.

Idempotency

AdCP v3 requires an idempotency_key on every mutating request. For sellers, that's create_media_buy, update_media_buy, sync_creatives, and any sync_* tools you implement. Idempotency is wired in the Implementation example above — this section explains what the framework does for you and the subtleties to know.

What the framework handles when you pass idempotency to createAdcpServerFromPlatform:

• Rejects missing or malformed idempotency_key with INVALID_REQUEST. The spec pattern is ^[A-Za-z0-9_.:-]{16,255}$ — a test key like "key1" will be rejected for length, not idempotency logic.
• Hashes the request payload with RFC 8785 JCS; returns IDEMPOTENCY_CONFLICT on same-key-different-payload. The error body carries only code + message — no payload hash, no field pointer, no leaked cached content.
• Returns IDEMPOTENCY_EXPIRED when a key is past the TTL (with ±60s clock-skew tolerance).
• Injects replayed: true on result.structuredContent.replayed when returning a cached response; fresh executions omit the field.
• Auto-declares adcp.idempotency.replay_ttl_seconds on get_adcp_capabilities.
• Only caches successful responses — errors re-execute on retry so transient failures don't lock into the cache.
• Atomic claim on check() so concurrent retries with a fresh key don't all race to execute side effects.

Scoping: the principal comes from resolveSessionKey (or override with resolveIdempotencyPrincipal(ctx, params, toolName) for per-tool custom scopes). Two callers with the same principal share a cache namespace; different principals are isolated.

Two things to know:

1. ttlSeconds must be 3600 (1h) to 604800 (7d) — out of range throws at createIdempotencyStore construction. Don't pass minutes thinking they're seconds.
2. If you register mutating handlers without passing idempotency, the framework logs an error at server-creation time (v3 non-compliance). Silence it by either wiring idempotency or setting capabilities.idempotency.replay_ttl_seconds in your config (declares non-compliance to buyers).

Going to Production

The quick-start uses memoryBackend() for idempotency and InMemoryStateStore for state — both reset on process restart and don't scale across replicas. Production swaps three pieces:

LEGACY (v5) — example below uses createAdcpServer. The Postgres wiring (pgBackend, PostgresStateStore, PostgresTaskStore) is identical for v6 — pass them to createAdcpServerFromPlatform(platform, { stateStore, taskStore, idempotency }) instead. For new agents, see the v6 worked example in skills/build-decisioning-platform/SKILL.md § Production wiring, or use the pool shortcut: createAdcpServerFromPlatform(platform, { pool }) auto-wires all three from a single pg.Pool.

import { Pool } from 'pg';
import {
  createIdempotencyStore,
  pgBackend,
  getIdempotencyMigration,
  PostgresStateStore,
  getAdcpStateMigration,
  PostgresTaskStore,
  MCP_TASKS_MIGRATION,
  cleanupExpiredIdempotency,
} from '@adcp/sdk/server';

// Fail fast — pg silently defaults to localhost+OS-user if DATABASE_URL is
// missing, which works on a dev laptop and breaks cryptically in CI.
if (!process.env.DATABASE_URL) {
  throw new Error('DATABASE_URL environment variable is required');
}
const pool = new Pool({ connectionString: process.env.DATABASE_URL });

// Run once per deployment before starting the server (e.g., as a
// separate migrate step, or at boot with a feature flag).
await pool.query(getIdempotencyMigration());
await pool.query(getAdcpStateMigration());
await pool.query(MCP_TASKS_MIGRATION);

const idempotency = createIdempotencyStore({
  backend: pgBackend(pool),
  ttlSeconds: 86400,
});
const stateStore = new PostgresStateStore(pool);
const taskStore = new PostgresTaskStore(pool);

// Cleanup expired idempotency rows hourly so the cache table doesn't
// grow unboundedly. Schedule via cron in production.
setInterval(() => cleanupExpiredIdempotency(pool).catch(console.error), 3600 * 1000);

serve(() =>
  createAdcpServer({
    name: 'My Seller Agent',
    version: '1.0.0',
    taskStore,
    stateStore,
    idempotency,

    // Real multi-tenant principal resolution — derived from an authenticated
    // session (e.g., JWT claims middleware before serve()), not a constant.
    resolveAccount: async ref => db.findAccount(ref),
    resolveSessionKey: ctx => (ctx.account as { id?: string } | undefined)?.id ?? 'unknown-principal',

    mediaBuy: {
      /* handlers */
    },
  })
);

Two things the example doesn't wire (app-specific):

• Authentication — the quick-start has no auth. Production agents need bearer-token or OAuth in front of serve(). The library provides OAuth helpers; bearer is middleware territory (Express/Fastify).
• Connection-pool sizing — pass max, idleTimeoutMillis, connectionTimeoutMillis on new Pool({...}) per your deployment's concurrency characteristics. The pg driver defaults are fine for low traffic.

Validation

After writing the agent, validate it. Fix failures. Repeat.

Full validation (if you can bind ports):

npx tsx agent.ts &
npx @adcp/sdk@latest storyboard run http://localhost:3001/mcp media_buy_seller --json

Sandbox validation (if ports are blocked):

npx tsc --noEmit

When storyboard output shows failures, fix each one:

• response_schema → response doesn't match Zod schema
• field_present → required field missing
• MCP error → check tool registration (schema, name)

Keep iterating until all steps pass.

Storyboards

Storyboard	Use case
media_buy_seller	Full lifecycle — every seller should pass this
media_buy_non_guaranteed	Auction flow with bid adjustment
media_buy_guaranteed_approval	IO approval workflow
media_buy_proposal_mode	AI-generated proposals
media_buy_catalog_creative	Catalog sync + conversions
schema_validation	Schema compliance + date validation errors
deterministic_testing	State machine correctness via comply_test_controller

Common Mistakes

Mistake	Fix
Using createTaskCapableServer + server.tool()	Use createAdcpServerFromPlatform(platform, opts) — handles schemas, response builders, capabilities, ctx_metadata
Calling createAdcpServer directly in new code	Reach for createAdcpServerFromPlatform first; createAdcpServer lives at @adcp/sdk/server/legacy/v5 for mid-migration / escape-hatch use only
Using module-level Maps for state	Use ctx.store — persists across HTTP requests, swappable for postgres
Return raw JSON without response builders	The framework auto-applies response builders — just return the data
Missing brand/operator in sync_accounts response	Echo them back from the request — they're required
sync_governance returns wrong shape	Must include status: 'synced' and governance_agents array
sandbox: false on mock data	Buyers may treat mock data as real
Returns raw JSON for validation failures	Use adcpError('INVALID_REQUEST', { message }) — storyboards validate the adcp_error structure
Missing publisher_properties or format_ids on Product	Both are required — see product example in get_products section
format_ids in products don't match list_creative_formats	Buyers echo format_ids from products into sync_creatives — if your validation rejects your own format_ids, the buyer can't fulfill creative requirements
Missing @types/node in devDependencies	process.env doesn't resolve without it — see Setup section
Dropping context from responses	Echo args.context back unchanged in every response — buyers use it for correlation
channels typed as string[] instead of MediaChannel[]	Use as const on channel arrays: channels: ['display', 'olv'] as const. TypeScript infers string[] from array literals, but the SDK requires the MediaChannel union type.

Reference

• docs/guides/BUILD-AN-AGENT.md — createAdcpServerFromPlatform patterns, async tools, state persistence
• docs/llms.txt — full protocol reference
• docs/TYPE-SUMMARY.md — curated type signatures
• storyboards/media_buy_seller.yaml — full buyer interaction sequence
• examples/error-compliant-server.ts — seller with error handling
• src/lib/server/create-adcp-server.ts — framework source (for TypeScript autocomplete exploration)