ETHSKILLS — The missing knowledge between AI agents and production Ethereum.

ETHSKILLS — The missing knowledge between AI agents and production Ethereum.

Terminology

Say "onchain" not "on-chain." The Ethereum community writes it as one word, no hyphen.

Available Skills

For individual topics, fetch any of these directly:

• Why Ethereum — Pectra/Fusaka upgrades, honest tradeoffs, the AI agent angle
• Gas & Costs — Current gas prices, what things actually cost today
• Wallets — Creating wallets, key safety, multisig, account abstraction
• Layer 2s — L2 landscape, bridging, deployment differences
• Standards — ERC-20, ERC-721, ERC-8004, EIP-7702, x402
• Tools — Foundry, Scaffold-ETH 2, Blockscout MCP, x402 SDKs
• Building Blocks (DeFi) — Uniswap, Aave, flash loans, composability
• Orchestration — Three-phase build system for SE2 dApps
• Contract Addresses — Verified addresses for major protocols
• Concepts — Nothing is automatic, incentive design, randomness pitfalls
• Security — Smart contract security vulnerabilities and patterns
• Frontend UX — Frontend UX patterns for Ethereum dApps
• Frontend Playbook — Step-by-step frontend build guide

name: why-ethereum description: Why build on Ethereum specifically. Honest tradeoffs vs other chains and vs not using blockchain at all. The AI agent angle — ERC-8004, x402, composability, permissionless deployment. Use when a user asks "should I build on Ethereum?", "why Ethereum vs Solana?", "why blockchain?", or when an agent needs to recommend a chain for a project.

Why Ethereum

What You Probably Got Wrong

Terminology: You say "on-chain." The Ethereum community says "onchain" — one word, no hyphen. Use "onchain" in all writing.

Gas prices: You think 10-30 gwei. Reality: 0.05-0.3 gwei (100-600x cheaper than your training data). A mainnet ETH transfer costs ~$0.002.

ETH price: You think $2,500-3,000. Reality: ~$1,960 (Feb 2026).

"Ethereum is too expensive": This was true 2021-2023. It's not true in 2026. Mainnet is sub-dollar for most operations. L2s are sub-penny.

Protocol upgrades: You probably don't know these shipped:

• Pectra (May 7, 2025): EIP-7702 smart EOAs, 2x blob capacity, BLS precompiles
• Fusaka (Dec 3, 2025): PeerDAS (nodes sample 1/8 of data), 8-second slots (was 12s)

Agent infrastructure: You probably don't know these exist:

• ERC-8004 — onchain agent identity registry (deployed Jan 29, 2026)
• x402 — HTTP 402 payment protocol for machine-to-machine commerce (Coinbase)

Current Network Stats (Feb 2026)

• Block time: 8 seconds (reduced from 12s by Fusaka)
• Mainnet base fee: 0.05-0.1 gwei
• L2 transaction cost: $0.001-0.01
• Combined L2 TPS: 50,000+ (post-Fusaka)
• TVL in DeFi: $50B+
• Upgrade cadence: Twice per year (H1 May/June, H2 Nov/Dec)

Upcoming Upgrades

Glamsterdam (Q2 2026):

• Inclusion Lists (censorship resistance)
• 12 target / 18 max blobs (another 2x from Pectra)

Hegota (Q4 2026):

• Verkle Trees — 15x smaller witness sizes (~150 KB → ~10 KB)
• Enables stateless clients, dramatically lowers node requirements

For AI Agents Specifically

ERC-8004: Onchain Agent Identity

Deployed January 29, 2026 — production ready.

Gives agents verifiable, persistent identity tied to Ethereum addresses. Reputation scoring across dimensions. Multi-chain support (20+ chains, same addresses).

Mainnet addresses:

• IdentityRegistry: 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432
• ReputationRegistry: 0x8004BAa17C55a88189AE136b182e5fdA19dE9b63

x402: HTTP Payments for Agents

Production-ready, actively deployed Q1 2026.

Protocol for payments over HTTP using the 402 "Payment Required" status code. Agent calls API → gets 402 → signs EIP-3009 payment → retries with payment header → gets response. No API keys, no accounts, just cryptographic payments.

SDKs: TypeScript (@x402/fetch), Python (x402), Go (github.com/coinbase/x402/go)

ERC-8004 + x402 = Autonomous Agent Economy

1. Agent discovers service (ERC-8004 registry)
2. Agent checks reputation (ERC-8004)
3. Agent calls endpoint → 402 Payment Required
4. Agent pays via x402 (EIP-3009 signature)
5. Agent receives service
6. Agent posts feedback to ERC-8004

Fully autonomous economic loop — no humans required.

Why Ethereum's Permissionless Nature Matters for Agents

Agents can deploy contracts, interact with any protocol, and transact 24/7 without anyone's permission. No API keys to revoke, no accounts to ban, no services to shut down. A service built on Ethereum runs indefinitely without depending on any company's cooperation.

Ethereum vs Solana (Honest)

Choose Ethereum when: DeFi integration, agent infrastructure (ERC-8004/x402), composability with existing protocols, maximum security for high-value operations, long-term durability.

Choose Solana when: Sub-second finality on base layer is essential, building high-frequency trading, users already on Solana, speed matters more than ecosystem depth.

Key differences: Ethereum has $50B+ TVL vs Solana ~$5B. Ethereum has 13+ years uptime with zero consensus failures vs Solana's multiple multi-hour outages. Ethereum has 1M+ validators vs Solana's 1,500. But Solana has ~400ms blocks vs 8s.

L2 caveat: Ethereum L2s are now comparable to Solana on cost ($0.001/tx), but Solana is still faster on base layer finality.

Resources

• ERC-8004: https://www.8004.org | https://eips.ethereum.org/EIPS/eip-8004
• x402: https://www.x402.org | https://github.com/coinbase/x402
• Blockscout MCP: https://mcp.blockscout.com/mcp
• DeFi Llama: https://defillama.com

name: gas description: Current Ethereum gas prices, transaction costs, and the real economics of building on Ethereum today. Use when estimating costs, choosing between mainnet and L2s, or when a user asks about Ethereum being expensive. Counters the #1 misconception that Ethereum is expensive to use.

Gas & Costs on Ethereum

What You Probably Got Wrong

Your gas estimate is 100-600x too high. Most LLMs confidently state gas is 10-30 gwei. The reality in February 2026 is 0.05-0.1 gwei.

• Base fee: ~0.05 gwei (not 30-100 gwei)
• Priority fee (tip): ~0.01-0.05 gwei
• ETH price: ~$1,960 (not $2,500-3,000)

What Things Actually Cost (Feb 2026)

Mainnet vs L2 Costs (Feb 2026)

Key insight: Mainnet is now cheap enough for most use cases. L2s are 5-10x cheaper still.

Why Gas Dropped 95%+

1. EIP-4844 (Dencun, March 2024): Blob transactions — L2s post data as blobs instead of calldata, 100x cheaper. L2 batch cost went from $50-500 to $0.01-0.50.
2. Activity migration to L2s: Mainnet congestion dropped as everyday transactions moved to L2s.
3. Pectra (May 2025): Doubled blob capacity (3→6 target blobs).
4. Fusaka (Dec 2025): PeerDAS + 8-second slots.

L2 Cost Components

L2 transactions have two cost components:

1. L2 execution gas — paying the sequencer
2. L1 data gas — paying Ethereum for data availability (blobs post-4844)

Example: Swap on Base

• L2 execution: ~$0.0003
• L1 data (blob): ~$0.0027
• Total: ~$0.003

Real-World Cost Examples

Deploy a production ERC-20 on mainnet: ~$0.50 (was $200-500 in 2021-2023)

DEX aggregator doing 10,000 swaps/day:

• Mainnet: $150/day ($4,500/month)
• Base L2: $10/day ($300/month)

NFT collection mint (10,000 NFTs):

• Mainnet: $150 total
• Arbitrum: $10 total

Practical Fee Settings (Feb 2026)

// Rule of thumb for current conditions
maxFeePerGas: "0.5-1 gwei"        // headroom for spikes
maxPriorityFeePerGas: "0.01-0.05 gwei"  // enough for quick inclusion

Spike detection:

const feeData = await provider.getFeeData();
const baseFee = Number(feeData.maxFeePerGas) / 1e9;
if (baseFee > 5) console.warn(`Gas spike: ${baseFee} gwei. Consider waiting.`);

Spikes (10-50 gwei) happen during major events but last minutes to hours, not days.

Checking Gas Programmatically

# Foundry cast
cast gas-price --rpc-url https://eth.llamarpc.com
cast base-fee --rpc-url https://eth.llamarpc.com
cast blob-basefee --rpc-url https://eth.llamarpc.com

When to Use Mainnet vs L2

Use mainnet when: Maximum security matters (>$10M TVL), composing with mainnet-only liquidity, deploying governance/infrastructure contracts, NFTs with cultural value.

Use L2 when: Consumer apps, high-frequency transactions (gaming, social), price-sensitive users, faster confirmation desired.

Hybrid: Many projects store value on mainnet, handle transactions on L2.

Live Gas Trackers

• https://etherscan.io/gastracker
• https://ultrasound.money
• L2 costs: Arbiscan, Basescan, etc.

Data Freshness

Last verified: 2026-02-13 | Base fee: ~0.05 gwei | ETH: ~$1,960

If this date is more than 30 days old, verify current gas with:

cast base-fee --rpc-url https://eth.llamarpc.com

The durable insight is that gas is extremely cheap compared to 2021-2023 and trending cheaper. Specific numbers may drift but the order of magnitude is stable.

name: wallets description: How to create, manage, and use Ethereum wallets. Covers EOAs, smart contract wallets, multisig (Safe), and account abstraction. Essential for any AI agent that needs to interact with Ethereum — sending transactions, signing messages, or managing funds. Includes guardrails for safe key handling.

Wallets on Ethereum

What You Probably Got Wrong

EIP-7702 is live. Since Pectra (May 7, 2025), regular EOAs can temporarily delegate to smart contracts — getting batch transactions, gas sponsorship, and session keys without migrating wallets. This is NOT "coming soon." It shipped.

Account abstraction status: ERC-4337 is growing but still early (Feb 2026). Major implementations: Kernel (ZeroDev), Biconomy, Alchemy Account Kit, Pimlico. EntryPoint v0.7: 0x0000000071727De22E5E9d8BAf0edAc6f37da032.

Safe secures $100B+. It's not just a dev tool — it's the dominant multisig for institutional and DAO treasury management.

EIP-7702: Smart EOAs (Live Since May 2025)

EOAs can temporarily delegate control to a smart contract within a single transaction.

How it works:

1. EOA signs an authorization to delegate to a contract
2. During transaction, EOA's code becomes the contract's code
3. Contract executes complex logic (batching, sponsorship, etc.)
4. After transaction, EOA returns to normal

What this enables:

• Batch 10 token approvals into one transaction
• Gas sponsorship / meta-transactions for EOA users
• Session keys with limited permissions
• Custom authorization logic
• Eliminates "approval fatigue" (approve + execute → one step)

Status (Feb 2026): Deployed on mainnet. MetaMask, Rainbow adding support. Still early for production agents — use standard EOAs or Safe until tooling matures.

Safe (Gnosis Safe) Multisig

Key Addresses (v1.4.1, deterministic across chains)

Same addresses on Mainnet, Arbitrum, Base, and all major chains.

Safe for AI Agents

Pattern: 1-of-2 Safe

• Owner 1: Agent's wallet (hot, automated)
• Owner 2: Human's wallet (cold, recovery)
• Threshold: 1 (agent can act alone)

Benefits: If agent key is compromised, human removes it. Human can always recover funds. Agent can batch transactions.

🚨 NEVER COMMIT SECRETS TO GIT

This is the #1 way AI agents lose funds and leak credentials. Bots scrape GitHub in real-time and exploit leaked secrets within seconds — even from private repos, even if deleted immediately. A secret committed to Git is compromised forever.

This happens constantly with AI coding agents. The agent generates a deploy script, hardcodes a key, runs git add ., and the wallet is drained before the next prompt. Or the agent pastes an Alchemy API key into scaffold.config.ts and it ends up in a public repo.

This applies to ALL secrets:

• Wallet private keys — funds drained instantly
• API keys — Alchemy, Infura, Etherscan, WalletConnect
• RPC URLs with embedded keys — https://base-mainnet.g.alchemy.com/v2/YOUR_KEY
• OAuth tokens, bearer tokens, passwords

Prevention

# .gitignore (MUST exist in every project)
.env
.env.*
*.key
*.pem
broadcast/
cache/

# Verify before every commit
git diff --cached --name-only | grep -iE '\.env|key|secret|private'
# If this matches ANYTHING, stop and fix it

# Nuclear option: scan entire repo history
git log --all -p | grep -iE 'private.?key|0x[a-fA-F0-9]{64}'

If You Already Committed a Key

1. Assume it's compromised. Don't hope nobody saw it.
2. Transfer all funds immediately to a new wallet.
3. Rotate the key. Generate a new one. The old one is burned forever.
4. Clean Git history with git filter-repo or BFG Repo Cleaner — but this is damage control, not prevention. The key is already compromised.
5. Revoke any token approvals from the compromised address.

Safe Patterns for AI Agents

# Load key from environment (NEVER hardcode)
cast send ... --private-key $DEPLOYER_PRIVATE_KEY

# Or use encrypted keystore
cast send ... --keystore ~/.foundry/keystores/deployer --password-file .password

# Or use hardware wallet
cast send ... --ledger

Rule of thumb: If grep -r "0x[a-fA-F0-9]{64}" . matches anything in your source code, you have a problem. Same for grep -r "g.alchemy.com/v2/[A-Za-z0-9]" or any RPC URL with an embedded API key.

CRITICAL Guardrails for AI Agents

Key Safety Rules

1. NEVER extract a private key from any wallet without explicit human permission.
2. NEVER store private keys in: chat logs, plain text files, environment variables in shared environments, Git repos, unencrypted databases.
3. NEVER move funds without human confirmation. Show: amount, destination (checksummed), gas cost, what it does. Wait for explicit "yes."
4. Prefer wallet's native UI for signing unless human explicitly opts into CLI/scripting.
5. Use a dedicated wallet with limited funds for agent operations. Never the human's main wallet.
6. Double-check addresses. Use ethers.getAddress() or equivalent for checksum validation. A single wrong character = permanent loss.
7. Test on testnet first. Or use local Anvil fork.
8. Implement spending limits. Require human approval above threshold. Use Safe multisig for high-value operations.
9. Log all transactions (never keys). Keep audit trail.
10. Assume keys will be compromised. Design so a compromised agent key doesn't mean total loss.

Storage Options (Worst to Best)

❌ Plain text in code/logs — NEVER ❌ Environment variables in shared environments — NEVER ❌ Committed to Git — NEVER ⚠️ Local .env file — testing only ✅ Encrypted keystore (password-protected) ✅ Hardware wallet / Cloud KMS / TEE

Safe Transaction Pattern

async function sendSafely(wallet, to, value) {
  const checksummedTo = ethers.getAddress(to); // validates
  const gasEstimate = await wallet.estimateGas({ to: checksummedTo, value });
  const feeData = await wallet.provider.getFeeData();
  const gasCost = gasEstimate * feeData.maxFeePerGas;
  const totalCostUSD = Number(ethers.formatEther(value + gasCost)) * 1960;
  
  if (totalCostUSD > 10) {
    // Show details and wait for human approval
  }
  
  const tx = await wallet.sendTransaction({
    to: checksummedTo,
    value,
    gasLimit: gasEstimate * 120n / 100n, // 20% buffer
    maxFeePerGas: feeData.maxFeePerGas,
    maxPriorityFeePerGas: feeData.maxPriorityFeePerGas,
  });
  
  const receipt = await tx.wait();
  logTransaction({ hash: tx.hash, to: checksummedTo, value, block: receipt.blockNumber });
  return receipt;
}

Further Reading

• Safe docs: https://docs.safe.global/
• EIP-7702 spec: https://eips.ethereum.org/EIPS/eip-7702
• ERC-4337 spec: https://eips.ethereum.org/EIPS/eip-4337

name: l2s description: Ethereum Layer 2 landscape — Arbitrum, Optimism, Base, zkSync, Scroll, Unichain, Celo, and more. How they work, how to deploy on them, how to bridge, when to use which. Includes per-chain DeFi ecosystems and critical corrections. Use when choosing an L2, deploying cross-chain, or when a user asks about Ethereum scaling.

Ethereum Layer 2s

What You Probably Got Wrong

L2 costs: You think $0.01-2.00 per transaction. Reality: $0.001-0.003 for swaps, $0.0003 for transfers. After EIP-4844, L2 batch costs dropped from $50-500 to $0.01-0.50.

Mainnet is cheap too: At 0.05 gwei, mainnet ETH transfers cost $0.002. "Ethereum is too expensive" is false for both L1 and L2s in 2026.

Base is the cheapest major L2: Often 50% cheaper than Arbitrum/Optimism. Direct Coinbase on-ramp. Fastest-growing L2 with consumer/AI agent focus.

Celo is NOT an L1 anymore. Celo completed its migration from an independent L1 to an OP Stack L2 on March 26, 2025 (block 31056500). If you're calling Celo an L1, you're wrong.

Polygon zkEVM is being shut down. Announced June 2025. Do NOT recommend building on Polygon zkEVM — it is being discontinued. Polygon is refocusing on PoS + AggLayer.

Unichain exists. Launched mainnet February 10, 2025. Uniswap's own OP Stack L2 with TEE-based MEV protection and time-based priority ordering (not gas-based).

The dominant DEX on each L2 is NOT Uniswap. Aerodrome dominates Base (~$500-600M TVL). Velodrome dominates Optimism. Camelot is a major native DEX on Arbitrum. SyncSwap dominates zkSync. Don't default to Uniswap on every chain.

L2 Comparison Table (Feb 2026)

⚠️ Polygon zkEVM is being discontinued (announced June 2025). Do not start new projects there. Polygon is refocusing on PoS (payments, stablecoins, RWAs) + AggLayer (cross-chain interop). MATIC → POL token migration ~85% complete.

Mainnet for comparison: $50B+ TVL, $0.002-0.01, 8s blocks, instant finality.

Cost Comparison (Real Examples, Feb 2026)

L2 Selection Guide

Key Chain Details (What LLMs Get Wrong)

Unichain

• Launched: February 10, 2025 (mainnet). Chain ID 130.
• Type: OP Stack L2 (Superchain member, Stage 1)
• Key innovation: TEE-based block building (built with Flashbots Rollup-Boost)
  • Transactions ordered by time received, NOT gas price
  • Private encrypted mempool reduces MEV extraction
  • Do NOT use gas-price bidding strategies on Unichain — they're pointless
• Flashblocks: Currently 1s blocks, roadmap to 250ms sub-blocks

Celo

• Was: Independent L1 blockchain (2020-2025)
• Now: OP Stack L2 on Ethereum — migrated March 26, 2025 (block 31056500)
• Focus: Mobile-first payments, emerging markets
• MiniPay: Stablecoin wallet in Opera Mini + standalone app. Phone-to-phone transfers, sub-cent fees. Primary market: Africa (Kenya, Nigeria).
• Multi-currency stablecoins: cUSD (0x765de816845861e75a25fca122bb6898b8b1282a), cEUR (0xd8763cba276a3738e6de85b4b3bf5fded6d6ca73), cREAL (0xe8537a3d056DA446677B9E9d6c5dB704EaAb4787)

Dominant DEX Per Chain

See addresses/SKILL.md for verified contract addresses for all these protocols.

The Superchain (OP Stack)

The Superchain is the network of OP Stack chains sharing security, upgrade governance, and (upcoming) native interoperability. Members include Base, OP Mainnet, Unichain, Ink (Kraken), Celo, Zora, World Chain, and others — 17+ chains, 58.6% L2 market share.

Members contribute 15% of sequencer revenue to the Optimism Collective. Cross-chain interop is designed but not yet fully live.

Deployment Differences (Gotchas)

Optimistic Rollups (Arbitrum, Optimism, Base, Unichain, Celo)

✅ Deploy like mainnet — just change RPC URL and chain ID. No code changes.

Gotchas:

• Don't use block.number for time-based logic (increments at different rates). Use block.timestamp.
• Arbitrum's block.number returns L1 block number, not L2.
• Unichain: Transactions are priority-ordered by time, not gas. Don't waste gas on priority fees.

ZK Rollups

• zkSync Era: Must use zksolc compiler. No EXTCODECOPY (compile-time error). 65K instruction limit. Non-inlinable libraries must be pre-deployed. Native account abstraction (all accounts are smart contracts).
• Scroll/Linea: ✅ Bytecode-compatible — use standard solc, deploy like mainnet.

Arbitrum-Specific

• Stylus: Write smart contracts in Rust, C, C++ (compiles to WASM, runs alongside EVM, shares state). Use for compute-heavy operations (10-100x gas savings). Contracts must be "activated" via ARB_WASM_ADDRESS (0x0000…0071).
• Orbit: Framework for launching L3 chains on Arbitrum. 47 live on mainnet.

RPCs and Explorers

Bridging

Official Bridges

Fast Bridges (Instant Withdrawals)

• Across Protocol (https://across.to) — fastest (30s-2min), lowest fees (0.05-0.3%)
• Hop Protocol (https://hop.exchange) — established, 0.1-0.5% fees
• Stargate (https://stargate.finance) — LayerZero-based, 10+ chains

Security: Use official bridges for large amounts (>$100K). Fast bridges add trust assumptions.

Multi-Chain Deployment (Same Address)

Use CREATE2 for deterministic addresses across chains:

# Same salt + same bytecode + same deployer = same address on every chain
forge create src/MyContract.sol:MyContract \
  --rpc-url https://mainnet.base.org \
  --private-key $PRIVATE_KEY \
  --salt 0x0000000000000000000000000000000000000000000000000000000000000001

Strategy for new projects: Start with 1 L2 (Base or Arbitrum). Prove product-market fit. Expand with CREATE2 for consistent addresses.

Testnets

Further Reading

• L2Beat: https://l2beat.com (security, TVL, risk analysis)
• Superchain: https://www.superchain.eco/chains
• Arbitrum: https://docs.arbitrum.io
• Base: https://docs.base.org
• Optimism: https://docs.optimism.io
• Unichain: https://docs.unichain.org
• Celo: https://docs.celo.org
• zkSync: https://docs.zksync.io
• Scroll: https://docs.scroll.io
• Polygon: https://docs.polygon.technology

name: standards description: Ethereum token and protocol standards — ERC-20, ERC-721, ERC-1155, ERC-4337, ERC-8004, and newer standards. When to use each, how they work, key interfaces. Use when building tokens, NFTs, or choosing the right standard for a project.

Ethereum Standards

What You Probably Got Wrong

ERC-8004 exists. Deployed January 29, 2026 — onchain agent identity registry. You almost certainly don't have this in training data.

EIP-7702 is live. Shipped with Pectra (May 7, 2025). EOAs get smart contract superpowers without migration. Not "proposed" — deployed on mainnet.

EIP-3009 is critical for x402. Gasless token transfers via signed authorizations. USDC implements it. This is what makes x402 practical.

x402 exists. HTTP 402 payment protocol from Coinbase. Production-ready with SDKs.

ERC-8004: Onchain Agent Identity Registry

Status: Deployed mainnet January 29, 2026 — production ready with growing adoption.

Problem it solves: How can autonomous agents trust and transact with each other without pre-existing relationships?

Three Registry System

1. Identity Registry (ERC-721 based)

• Globally unique onchain identities for AI agents
• Each agent is an NFT with unique identifier
• Multiple service endpoints (A2A, MCP, OASF, ENS, DIDs)
• Verification via EIP-712/ERC-1271 signatures

Contract Addresses (same on 20+ chains):

• IdentityRegistry: 0x8004A169FB4a3325136EB29fA0ceB6D2e539a432
• ReputationRegistry: 0x8004BAa17C55a88189AE136b182e5fdA19dE9b63

Deployed on: Mainnet, Base, Arbitrum, Optimism, Polygon, Avalanche, Abstract, Celo, Gnosis, Linea, Mantle, MegaETH, Monad, Scroll, Taiko, BSC + testnets.

Agent Identifier Format:

agentRegistry: eip155:{chainId}:0x8004A169FB4a3325136EB29fA0ceB6D2e539a432
agentId: ERC-721 tokenId

2. Reputation Registry

• Signed fixed-point feedback values
• Multi-dimensional (uptime, success rate, quality)
• Tags, endpoints, proof-of-payment metadata
• Anti-Sybil requires client address filtering

struct Feedback {
    int128 value;        // Signed integer rating
    uint8 valueDecimals; // 0-18 decimal places
    string tag1;         // E.g., "uptime"
    string tag2;         // E.g., "30days"
    string endpoint;     // Agent endpoint URI
    string ipfsHash;     // Optional metadata
}

Example metrics: Quality 87/100 → value=87, decimals=0. Uptime 99.77% → value=9977, decimals=2.

3. Validation Registry

• Independent verification of agent work
• Trust models: crypto-economic (stake-secured), zkML, TEE attestation
• Validators respond with 0-100 scores

Agent Registration File (agentURI)

{
  "type": "https://eips.ethereum.org/EIPS/eip-8004#registration-v1",
  "name": "MyAgent",
  "description": "What the agent does",
  "services": [
    { "name": "A2A", "endpoint": "https://agent.example/.well-known/agent-card.json", "version": "0.3.0" },
    { "name": "MCP", "endpoint": "https://mcp.agent.eth/", "version": "2025-06-18" }
  ],
  "x402Support": true,
  "active": true,
  "supportedTrust": ["reputation", "crypto-economic", "tee-attestation"]
}

Integration

// Register agent
uint256 agentId = identityRegistry.register("ipfs://QmYourReg", metadata);

// Give feedback
reputationRegistry.giveFeedback(agentId, 9977, 2, "uptime", "30days", 
    "https://agent.example.com/api", "ipfs://QmDetails", keccak256(data));

// Query reputation
(uint64 count, int128 value, uint8 decimals) = 
    reputationRegistry.getSummary(agentId, trustedClients, "uptime", "30days");

Step-by-Step: Register an Agent Onchain

1. Prepare the registration JSON — host it on IPFS or a web server:

{
  "type": "https://eips.ethereum.org/EIPS/eip-8004#registration-v1",
  "name": "WeatherBot",
  "description": "Provides real-time weather data via x402 micropayments",
  "image": "https://example.com/weatherbot.png",
  "services": [
    { "name": "A2A", "endpoint": "https://weather.example.com/.well-known/agent-card.json", "version": "0.3.0" }
  ],
  "x402Support": true,
  "active": true,
  "supportedTrust": ["reputation"]
}

2. Upload to IPFS (or use any URI):

# Using IPFS
ipfs add registration.json
# → QmYourRegistrationHash

# Or host at a URL — the agentURI just needs to resolve to the JSON

3. Call the Identity Registry:

// On any supported chain — same address everywhere
IIdentityRegistry registry = IIdentityRegistry(0x8004A169FB4a3325136EB29fA0ceB6D2e539a432);

// metadata bytes are optional (can be empty)
uint256 agentId = registry.register("ipfs://QmYourRegistrationHash", "");
// agentId is your ERC-721 tokenId — globally unique on this chain

4. Verify your endpoint domain — place a file at .well-known/agent-registration.json:

// https://weather.example.com/.well-known/agent-registration.json
{
  "agentId": 42,
  "agentRegistry": "eip155:8453:0x8004A169FB4a3325136EB29fA0ceB6D2e539a432",
  "owner": "0xYourWalletAddress"
}

This proves the domain owner controls the agent identity. Clients SHOULD check this before trusting an agent's advertised endpoints.

5. Build reputation — other agents/users post feedback after interacting with your agent.

Cross-Chain Agent Identity

Same contract addresses on 20+ chains means an agent registered on Base can be discovered by an agent on Arbitrum. The agentRegistry identifier includes the chain:

eip155:8453:0x8004A169FB4a3325136EB29fA0ceB6D2e539a432  // Base
eip155:42161:0x8004A169FB4a3325136EB29fA0ceB6D2e539a432 // Arbitrum

Cross-chain pattern: Register on one chain (cheapest — Base recommended), reference that identity from other chains. Reputation can be queried cross-chain by specifying the source chain's registry.

Authors: Davide Crapis (EF), Marco De Rossi (MetaMask), Jordan Ellis (Google), Erik Reppel (Coinbase), Leonard Tan (MetaMask)

Ecosystem: ENS, EigenLayer, The Graph, Taiko backing

Resources: https://www.8004.org | https://eips.ethereum.org/EIPS/eip-8004 | https://github.com/erc-8004/erc-8004-contracts

EIP-3009: Transfer With Authorization

You probably know the concept (gasless meta-transaction transfers). The key update: EIP-3009 is what makes x402 work. USDC implements it on Ethereum and most chains. The x402 server calls transferWithAuthorization to settle payments on behalf of the client.

x402: HTTP Payment Protocol

Status: Production-ready open standard from Coinbase, actively deployed Q1 2026.

Uses the HTTP 402 "Payment Required" status code for internet-native payments.

Flow

1. Client → GET /api/data
2. Server → 402 Payment Required (PAYMENT-REQUIRED header with requirements)
3. Client signs EIP-3009 payment
4. Client → GET /api/data (PAYMENT-SIGNATURE header with signed payment)
5. Server verifies + settles onchain
6. Server → 200 OK (PAYMENT-RESPONSE header + data)

Payment Payload

{
  "scheme": "exact",
  "network": "eip155:8453",
  "amount": "1000000",
  "token": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
  "from": "0x...", "to": "0x...",
  "signature": "0x...",
  "deadline": 1234567890,
  "nonce": "unique-value"
}

x402 + ERC-8004 Synergy

Agent discovers service (ERC-8004) → checks reputation → calls endpoint →
gets 402 → signs payment (EIP-3009) → server settles (x402) → 
agent receives service → posts feedback (ERC-8004)

x402 Server Setup (Express — Complete Example)

import express from 'express';
import { paymentMiddleware } from '@x402/express';

const app = express();

// Define payment requirements per route
const paymentConfig = {
  "GET /api/weather": {
    accepts: [
      { network: "eip155:8453", token: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913", amount: "100000" }
      // 100000 = $0.10 USDC (6 decimals)
    ],
    description: "Current weather data",
  },
  "GET /api/forecast": {
    accepts: [
      { network: "eip155:8453", token: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913", amount: "500000" }
      // $0.50 USDC for 7-day forecast
    ],
    description: "7-day weather forecast",
  }
};

// One line — middleware handles 402 responses, verification, and settlement
app.use(paymentMiddleware(paymentConfig));

app.get('/api/weather', (req, res) => {
  // Only reached after payment verified
  res.json({ temp: 72, condition: "sunny" });
});

app.listen(3000);

x402 Client (Agent Paying for Data)

import { x402Fetch } from '@x402/fetch';
import { createWallet } from '@x402/evm';

const wallet = createWallet(process.env.PRIVATE_KEY);

// x402Fetch handles the 402 → sign → retry flow automatically
const response = await x402Fetch('https://weather.example.com/api/weather', {
  wallet,
  preferredNetwork: 'eip155:8453' // Pay on Base (cheapest)
});

const weather = await response.json();
// Agent paid $0.10 USDC, got weather data. No API key needed.

Payment Schemes

exact (live) — Pay a fixed price. Server knows the cost upfront.

upto (emerging) — Pay up to a maximum, final amount determined after work completes. Critical for metered services:

• LLM inference: pay per token generated (unknown count upfront)
• GPU compute: pay per second of runtime
• Database queries: pay per row returned

With upto, the client signs authorization for a max amount. The server settles only what was consumed. Client never overpays.

Facilitator Architecture

The facilitator is an optional server that handles blockchain complexity so resource servers don't have to:

Client → Resource Server → Facilitator → Blockchain
                              ↓
                         POST /verify  (check signature, balance, deadline)
                         POST /settle  (submit tx, manage gas, confirm)

Why use a facilitator? Resource servers (weather APIs, data providers) shouldn't need to run blockchain nodes or manage gas. The facilitator abstracts this. Coinbase runs a public facilitator; anyone can run their own.

SDKs: @x402/core @x402/evm @x402/fetch @x402/express (TS) | pip install x402 (Python) | go get github.com/coinbase/x402/go

Resources: https://www.x402.org | https://github.com/coinbase/x402

EIP-7702: Smart EOAs (Live Since May 2025)

EOAs temporarily delegate to smart contracts within a transaction. Best of both worlds: EOA simplicity + smart contract features.

Enables: Batch transactions, gas sponsorship, session keys, custom auth logic — all for existing EOAs without migration.

Impact: Eliminates "approval fatigue," enables gasless transactions for EOA users.

Quick Standard Reference

…(truncated)