菜单
Mainnet fork testing for smart contracts. Tests against real on-chain state at a pinned block number.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | fork-test |
| description | Mainnet fork testing for smart contracts. Tests against real on-chain state at a pinned block number. |
| triggers | ["test against mainnet","fork test","test with real state","simulate on mainnet","fork testing","test on a fork","mainnet simulation"] |
| agent | orchestrator |
| model | sonnet |
Writes and runs fork tests that validate contract behavior against real mainnet state. Always pins a specific block number for reproducibility. Uses debugger if any test fails.
eth_skill_start({ skill: "fork-test", totalPhases: 1 })integration-test-engineer with the target chain and contracts to testintegration-test-engineer selects a fork block, writes fork tests, and runs themdebugger with the failing test + revert datadebugger provides root cause → orchestrator spawns integration-test-engineer to fixeth_reset()"run fork tests against mainnet"
"test the Vault against Aave on a fork"
"simulate the swap on a mainnet fork"
# Output:
- test/fork/[ContractName].fork.t.sol
- trace logs for any failures
- Fork block number and chain documented in test file
ETH_RPC_URL (or chain-specific RPC var) in .envETH_RPC_URL. Add to .env."End-to-end protocol build pipeline — contract development, testing, auditing, gas optimization, deployment, and dApp integration. Hybrid sequential/parallel execution.
Shows all available eth-agents skills with trigger phrases and usage examples.
Extracts a reusable skill pattern from the current session and saves it to user scope for automatic activation in future sessions.
Security vulnerability analysis for Solidity smart contracts. Runs static analysis, optional dynamic testing and formal verification.
Full contract build pipeline — write → TDD loop → static audit → gas optimization. Produces a production-ready contract.
Multichain smart contract deployment with Etherscan verification. Requires explicit user confirmation before mainnet execution.