{ "markdown": "\"image\"\n\nLLMs think gas is 30 gwei. It's 0.05. They've never heard of ERC-8004 or x402. They hallucinate contract addresses. This site fixes that.\n\n## How It Works\n\nEach skill is a markdown file. Give any URL to your AI agent — it reads it and instantly corrects its Ethereum knowledge.\n\n```\nhttps://ethskills.com/SKILL.md ← table of contents\nhttps://ethskills.com/gas/SKILL.md ← just gas & costs\nhttps://ethskills.com/why/SKILL.md ← why ethereum\nhttps://ethskills.com/crops/SKILL.md ← CROPS architecture review\n```\n\n### Add the top-level SKILL.md to your agent's context\nThe agent will look up the specific skills when needed.\n\nJust prompt your agent with this:\n\n```\nInstall the https://github.com/austintgriffith/ethskills/blob/master/SKILL.md skill\n```\n\n### Claude Code Plugin\n\nInstall directly from GitHub — no clone required:\n\n```bash\n/plugin marketplace add austintgriffith/ethskills\n/plugin install ethskills@ethskills\n```\n\n## Terminology\n\n**Say \"onchain\" not \"on-chain.\"** The Ethereum community writes it as one word — no hyphen. LLMs default to \"on-chain\" but that's not how builders talk. Use \"onchain\" in all writing, documentation, and generated content.\n\n## Skills\n\n| Skill | What LLMs Get Wrong |\n|-------|-------------------|\n| [Why Ethereum](why/SKILL.md) | Pectra/Fusaka shipped, gas is cheap now, AI agent angle |\n| [Ship](ship/SKILL.md) | Start coding before planning architecture, over-engineer with 5+ contracts when 1-2 will do, skip chain selection and the CROPS gate |\n| [CROPS Review](crops/SKILL.md) | Miss custody, privacy, exit paths, censorship resistance, and centralization risks in otherwise working apps/contracts |\n| [Protocol](protocol/SKILL.md) | Confuse roadmap diagrams with shipped EIPs, cite shipped features (PeerDAS) as upcoming and deprioritized ones (Verkle) as still planned, miss the client-team decision process |\n| [Gas & Costs](gas/SKILL.md) | Think gas is 10-30 gwei — reality is 0.05-0.3 gwei |\n| [Wallets](wallets/SKILL.md) | EIP-7702 is live, Safe addresses, agent key safety |\n| [Layer 2s](l2s/SKILL.md) | Think L2 txs cost $0.01-2.00 — reality is <$0.001 |\n| [Standards](standards/SKILL.md) | Don't know ERC-8004, EIP-7702 status, EIP-3009 for x402 |\n| [Tools](tools/SKILL.md) | Don't know x402, Blockscout MCP, current tool landscape |\n| [Money Legos](building-blocks/SKILL.md) | Stale on current DeFi state, Uniswap V4 status |\n| [Orchestration](orchestration/SKILL.md) | Don't know SE2 three-phase build system |\n| [Contract Addresses](addresses/SKILL.md) | Hallucinate addresses — these are verified onchain |\n| [Concepts](concepts/SKILL.md) | Nothing is automatic, incentive design, randomness pitfalls |\n| [Security](security/SKILL.md) | Token decimals, reentrancy, oracle manipulation, vault inflation, pre-deploy checklist |\n| [Testing](testing/SKILL.md) | Test getters and OpenZeppelin internals, skip fuzz and fork tests on integrations, mock external protocols instead of forking |\n| [Indexing](indexing/SKILL.md) | Loop over blocks to read history, try to query historical state via raw RPC, ignore events as the primary read API |\n| [Frontend UX](frontend-ux/SKILL.md) | Onchain button rules, three-button approval flow, Address components, USD values |\n| [Frontend Playbook](frontend-playbook/SKILL.md) | Fork mode, IPFS deploy, Vercel config, ENS setup, production checklist |\n| [QA](qa/SKILL.md) | Treat \"deploys without errors\" as shipping, leave SE2 default branding, miss approve-button double-fire, USD values, mobile deep links |\n| [Audit](audit/SKILL.md) | Call one generic review pass an audit, miss systematic checklist coverage across AMM/lending/oracle/proxy/governance domains |\n\n## Security Guardrails\n\nSkills teach restraint, not just capability. Every skill that touches keys, credentials, or funds includes explicit safety rules — because LLMs optimize for speed and will hardcode a private key into `git add .` if you let them. Coverage includes wallet keys, API keys, RPC URLs, and the common SE2 `scaffold.config.ts` trap.\n\n## Methodology\n\nWe test stock LLMs, find what they get wrong, and write corrections. Content is verified against onchain reality. If an LLM already knows something, we don't include it.\n\n**Every proposed change goes through [triage](https://github.com/austintgriffith/ethskills-research/blob/master/research/triage-methodology.md):** spawn a stock LLM, give it a realistic task, see what it gets wrong. Only verified blind spots survive.\n\nSee the [research repo](https://github.com/austintgriffith/ethskills-research) for baseline audits, gap analysis, and full methodology.\n\n## Contributing\n\nSomething wrong or missing? Humans and agents are welcome to [open a PR](https://github.com/austintgriffith/ethskills/pulls). Read [CONTRIBUTING.md](CONTRIBUTING.md) first — the bar is \"would a stock LLM get this wrong?\"\n\n## License\n\nMIT\n", "bytes": 4990, "sha": "d0a1fa64c68f078c08d32447eccb215ebbadb1a86407884777659a718cbd0c69", "repo_slug": "austintgriffith/ethskills", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/skl_austintgriffith_ethskills_ethskills_f281f73d/readme" }