smart402 Blog Writing about AI agent infrastructure and x402 payments. https://www.smart402.com/blog 2026-04-01T00:00:00Z smart402 https://www.smart402.com smart402 0.4.0: AI Agent Spending Controls That Count What Actually Landed https://www.smart402.com/blog/smart402-0-4-0-ai-agent-spending-controls 2026-04-01T00:00:00Z 2026-04-01T00:00:00Z smart402 smart402 0.4.0 closes the loop on AI agent spending controls. Budget windows now count confirmed on-chain payments only — failed transactions return their allowance automatically. Dogfooding smart402 with AgentKit and x402 v2 https://www.smart402.com/blog/x402-base-mainnet-dogfood 2026-03-28T00:00:00Z 2026-03-28T00:00:00Z smart402 We ran smart402 against Base mainnet with a live Bazaar service and actual USDC. Here's what we set out to build, the wall we hit, and what we fixed along the way. smart402 0.2.0: Managed Wallets, AgentKit, and Why We're Excited https://www.smart402.com/blog/agentkit-managed-wallet-support 2026-03-23T00:00:00Z 2026-03-23T00:00:00Z smart402 The first feature request after launch: AgentKit compatibility. Smart402Guard no longer requires a signer object. Pass wallet_address as a plain string — works with AgentKit, Privy, Turnkey, or any EVM wallet. The Safest x402 Risk Layer Is One That Doesn't Speak Agent https://www.smart402.com/blog/no-llm-in-the-decision-path 2026-03-17T00:00:00Z 2026-03-19T00:00:00Z smart402 What happens after a prompt injection attack succeeds? smart402 sits outside the blast radius entirely: deterministic policy checks, no LLM, no attack surface that text can reach.