AI Orderability (AIO) Agents CPG with x402: Retail & Marketplace Sourcing
CPG sourcing for AI Agents with x402: BPC brands, agent IM, Trading Desk. GreenCore Solutions Corp.
使うべきか
品質と安全性
検出事項(3)
- HIGH
- MEDIUMget_x402_rail 内
- INFOget_brand_identifiers 内
ツール定義とプロトコルへの準拠に関する自動分析に基づいています。
コンテキストコスト
これは、サーバーのツールがモデルのコンテキストに読み込まれるたびに消費されるおおよそのトークン数です。数が多いほど、ほかのタスクに使える注意が減ります。
インストール
ワンクリックインストール
これを `claude_desktop_config.json` ファイルに追加してください:
{
"mcpServers": {
"gsc-marketplace": {
"url": "https://mcp.gsc-marketplace.ai/mcp"
}
}
}リモートエンドポイント
https://mcp.gsc-marketplace.ai/mcpstreamable-httpできること
ツール一覧
ツール(11)
🟢search_brands(market, query, limit)
When a marketplace or retail sourcing agent calls, it leaves with a brand it can act on — found, verified, and answerable — with Instant Messaging for agents or the human-in-the-loop GSC Trading Desk as next steps. Step 1 — FOUND: search the brands the CPG Knowledge Graph covers in one of 50 markets (live). Returns brand names from the graph's SPARKS depth surface; then call get_brand to verify, get_brand_identifiers for the identifier policy. Depth: 24,126 brands · 5,928 makers · 2,020 banners · 50 markets. Brands, makers and banners are separate registries — never summed.
入力スキーマ
{
"type": "object",
"properties": {
"market": {
"type": "string",
"minLength": 2,
"maxLength": 16,
"description": "SM-ECO-10060 member code or namespace, e.g. BR or BR-ECO-10060"
},
"query": {
"type": "string",
"maxLength": 80,
"description": "Optional name fragment (accent-insensitive)"
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 200,
"description": "Max results, default 50"
}
},
"required": [
"market"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_brand(brand, market)
Step 2 — VERIFIED: is the brand on the CPG Knowledge Graph for this market, and at what SPARKS completeness tier (100 = resolvable pack/size record held · 8 = registered, pending · floor = STANDARD default). Returns the market's supply/demand context and both doors to act. No prices, no stock, no terms — those go through the doors.
入力スキーマ
{
"type": "object",
"properties": {
"brand": {
"type": "string",
"minLength": 1,
"maxLength": 120,
"description": "Brand name as spelled on the graph"
},
"market": {
"type": "string",
"minLength": 2,
"maxLength": 16,
"description": "SM-ECO-10060 member code or namespace, e.g. BR or BR-ECO-10060"
}
},
"required": [
"brand",
"market"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_brand_identifiers(brand, market)
Step 3 — ANSWERABLE: GSC publishes ONLY its own GS1-registered identifiers (the 990832300xxx block) on any surface; third-party GTINs are never stored on the graph's entity universe and never appear here. A GTIN the caller already holds resolves live on the CPG Knowledge Graph (resolve_gtin, resolve_sparks, check_eligibility). The demo brand carries GSC's rail GTIN.
入力スキーマ
{
"type": "object",
"properties": {
"brand": {
"type": "string",
"minLength": 1,
"maxLength": 120,
"description": "Brand name"
},
"market": {
"type": "string",
"minLength": 2,
"maxLength": 16,
"description": "SM-ECO-10060 member code or namespace, e.g. BR or BR-ECO-10060"
}
},
"required": [
"brand"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_marketplace_operators(market)
The demand side per market as the CPG Knowledge Graph publishes it (retail banners — public registry data), each carrying an agentic-posture field from a fixed taxonomy and sorted by it. Neutral by design: no tiers, no pursuit language, no relationship claims, no named individuals. Posture is wire evidence only; 'unassessed' means none collected yet.
入力スキーマ
{
"type": "object",
"properties": {
"market": {
"type": "string",
"minLength": 2,
"maxLength": 16,
"description": "Omit for the 50-market summary"
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟡instant_message_agents
How an agent lodges an RFQ, a terms request, or an escalation with GreenCore Solutions Corp. — typed, identified, ticketed, human-signed (IA-MESSAGE). Points at the existing door: the x-gsc-inbound address and the transaction MCP. No new write path here.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}⚪contact_trading_desk
The human channel. The GSC Trading Desk is informed by Navigator across all GSC agentic assets — a human reads, a human answers. Use it for anything a person should decide; agents use Instant Messaging.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_x402_rail
What GSC-Marketplace declares about x402: header, network, asset, status — and where the single paid endpoint lives (https://gsc-marketplace.ai/api; its terms are stated in its 402 payload only, never here). Door: https://gsc-marketplace.ai/x402 · skill: x402-settlement.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_radar
The public network figures as GSC Radar publishes them (live JSON), macro numbers only.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_cpg_knowledge_graph
24,126 brands · 5,928 makers · 2,020 banners · 50 markets. The Beauty & Personal Care source of truth for AI agents, classified on SPARKS; 14 read tools over MCP; the data surface every marketplace answer resolves against.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_acm_68000
The deterministic signal protocol every GSC surface speaks: seven signals, GS1-anchored, append-only, non-substitutable. Canon resolves on acm-68000.ai / .org and the standards beacon; the gen-2 door is cpg-68000.ai.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_aio_agent_fleet
The AIO Agent class and its regional hosts: signed, resident, addressable resolvers with market × category remits — the NextGen identity register is the signed Card, verifiable against each host's jwks. Status is only what the wire says.
入力スキーマ
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}コミュニティ
エビデンス