Validate
AI-assistedAI helps select a bounded resource; the human approves the experiment.
Builds and tests a tiny x402-protected resource on Base Sepolia or Solana Devnet; it never uses mainnet or counts testnet settlement as revenue.
User-selected scenario, not an expected outcome.
Current level 2/5. Potential level 3/5 in 3–6 months.
AI helps select a bounded resource; the human approves the experiment.
AI can implement sandbox code under review.
AI prepares discovery copy; no live sale is attempted.
A testnet request runs only after explicit approval.
The human reviews receipts, security, and any future mainnet decision.
Approval applies to the exact action once. It is not blanket permission.
Approve each faucet-funded testnet request before it runs.
Approve any sandbox deployment target and rollback plan.
Approve public documentation; label all settlement and balances as testnet.
Do not make real purchases in this launch workflow.
Do not transfer real funds or assets.
Do not place trades or publish trading instructions.
Do not open a financial account.
Do not connect a live wallet, exchange, or financial credential.
Do not use mainnet or activate real payments.
/goal Goal: build a tiny x402-protected API experiment using only Base Sepolia or Solana Devnet, an ephemeral test signer, and faucet assets. Real-money budget is $0. Do not connect live wallet, exchange, brokerage, banking, or card credentials. Do not use mainnet, buy assets, transfer real value, trade, open accounts, or activate live payments. Choose one deterministic, low-risk digital resource such as a small public-data transformation or generated test fixture. Define its request, response, price in testnet units, rate limit, error contract, and evidence log. Implement the free discovery response and paid-resource gate in a local or explicitly approved sandbox. Add tests for missing payment, invalid signature, wrong network, expired authorization, replay, duplicate settlement, facilitator error, rate limit, and successful testnet response. Before any testnet transaction, show the network, faucet amount, recipient, resource, and expected result and wait for explicit approval. Record the testnet receipt as sandbox evidence, never revenue. Produce a real-sale readiness packet covering security review, monitoring, support, legal/compliance questions, owner approval, and rollback; do not execute that future phase. MILESTONES 1. Bound the resource: Request, response, testnet price, limits, and failure contract are approved. 2. Implement sandbox: Only Base Sepolia or Solana Devnet and ephemeral test material are configured. 3. Test safety: Authorization, replay, settlement, rate-limit, and error tests pass. 4. Record approved testnet proof: Human-approved receipt is labeled sandbox, not revenue. 5. Prepare future decision: Readiness packet exists; no live action occurs. COMPLETION CRITERIA 1. Testnet-only configuration 2. No live credentials or real expenditure 3. Safety tests pass 4. Testnet receipt is not revenue 5. Future live phase remains unexecuted BUDGET CEILING $0 USD EXTERNAL-ACTION GATES 1. Testnet transaction — Explicit approval: Approve each faucet-funded testnet request before it runs. 2. Deploy — Explicit approval: Approve any sandbox deployment target and rollback plan. 3. Publish — Explicit approval: Approve public documentation; label all settlement and balances as testnet. 4. Purchase — Prohibited: Do not make real purchases in this launch workflow. 5. Transfer — Prohibited: Do not transfer real funds or assets. 6. Trade — Prohibited: Do not place trades or publish trading instructions. 7. Open financial account — Prohibited: Do not open a financial account. 8. Connect credentials — Prohibited: Do not connect a live wallet, exchange, or financial credential. 9. Activate live payments — Prohibited: Do not use mainnet or activate real payments. STOP CONDITIONS 1. Wrong network 2. Real asset or credential requested 3. Safety test failure 4. Any real spend
Projected scenarios never appear here as observed outcomes.
The runnable asset is intentionally Draft; no sandbox execution is claimed.
No model test date recorded
Catalog baseline; exact version must be recorded at first test
A developer exploring programmatic API monetization without real-money exposure. · Prove one paid-resource flow on testnet and prepare a separately approved path toward a future first sale.