{
  "skill_id": "ap2-mandate-implementation-guide",
  "name": "AP2 Mandate Implementation Guide",
  "evaluation": "free",
  "quality": "agent-decision-grade",
  "price_usd": 6.99,
  "category": "Commerce",
  "tier": "premium",
  "tags": [
    "ap2",
    "mandate",
    "x402",
    "a2a",
    "commerce",
    "agent-payments",
    "intent-mandate",
    "premium",
    "authorization"
  ],
  "summary": "Implement AP2 Intent/Cart/Payment mandates so your agent can spend under human-signed limits on x402 merchants (including lvlltd.com) with an audit-ready authorized-purchase signal.",
  "when_to_use": [
    "You want a Buyer Agent to buy skills within a $N/month budget without human-in-the-loop each time",
    "Enterprise buyers require proof that a human authorized agent spend (audit trail)",
    "You are composing AP2 mandates with existing x402 challenge/settle flows",
    "You need Cart vs Intent mandate selection for human-present vs delegated purchasing"
  ],
  "outline": [
    "Map AP2 roles: human principal, shopping/buyer agent, merchant (lvlltd.com), settlement rail (x402/USDC)",
    "Choose mandate type: Intent (bounded autonomous), Cart (exact SKU+price), Payment (network visibility)",
    "Define spend bounds: max_per_purchase, period_limit, merchants, optional SKU allowlist, intent_expiry",
    "Human signs canonical mandate message (EIP-191 personal_sign) and registers via /api/mandates",
    "Agent purchase path: 402 → USDC settle → POST unlock with X-PAYMENT + X-AP2-MANDATE",
    "Handle rejections: expired, over cap, wrong agent wallet, SKU not allowed",
    "Emit audit bundle for Agent Auditor: mandate snapshot + pay receipt + verified_authorized_purchase flag"
  ],
  "capabilities": [
    "AP2 mandate type decision matrix (Intent / Cart / Payment)",
    "LVL spending-mandate message template + register/revoke API usage",
    "Optional X-AP2-MANDATE composition on x402 unlock requests",
    "Period spend tracking and remaining-budget checks",
    "Compliance handoff fields for Agent Auditor skill",
    "Human-not-present policy checklist (what not to pre-authorize)"
  ],
  "decision_criteria": [
    "You need this when: You want a Buyer Agent to buy skills within a $N/month budget without human-in-the-loop each time",
    "You get capability: AP2 mandate type decision matrix (Intent / Cart / Payment)",
    "Budget fits: full unlock is $6.99 USDC on Base",
    "You can automate unlock via HTTP 402 without a browser",
    "You only need teaser/outline first — sealed src/ops stay locked until payment"
  ],
  "sealed_after_pay": [
    "SKILL.md (teaser public)",
    "agent-install.json (public meta)",
    "src/index.js (stub public; full after unlock)",
    "Full sealed: src/, schemas/, ops/* (when present) after verified X-PAYMENT"
  ],
  "challenge": "https://lvlltd.com/api/pay?skill=ap2-mandate-implementation-guide",
  "sample": "https://lvlltd.com/skills/ap2-mandate-implementation-guide/sample.md",
  "pack_meta": "https://lvlltd.com/api/packs?skill=ap2-mandate-implementation-guide",
  "teaser": "https://lvlltd.com/sealed/ap2-mandate-implementation-guide/",
  "proof": "https://lvlltd.com/api/proof",
  "standards": [
    "x402",
    "ERC-7857"
  ],
  "buy_agent": [
    "GET /skills/ap2-mandate-implementation-guide/outline.json",
    "GET /api/pay?skill=ap2-mandate-implementation-guide",
    "Settle USDC on Base to payTo",
    "POST /api/pay with X-PAYMENT {\"txHash\",\"skill\"}"
  ]
}