This site is one window onto the launchpad, not the launchpad itself. Every DAO lives on-chain, permissionless to read and to trade — so anyone can ship their own explorer, or a dedicated site for their own DAO. These agent skills hand your coding agent everything it needs to build one.
An agent skill is a markdown briefing that a coding agent — Claude Code, Cursor, or any tool that reads project skills — loads before it starts building. Ours carry the deployed contract addresses, links to the hosted ABIs, the decimal conventions, and the sharp edges an agent would otherwise trip on (this chain’s modified Uniswap router being the big one — stock Uniswap SDKs encode swaps that fail here).
The flow: download a skill into your project, open your agent, and describe the interface you want. The skill steers the agent to read everything from public RPC — no API keys, no backend, no permission from ROHME — and to verify against mainnet with read-only calls before calling it shipped.
Build an interface for the whole launchpad — enumerate every DAO from the factory, render live prices and backing through the lens, swap on the v4 curves, and found new DAOs. Everything reads from public RPC; no API keys.
mkdir -p .claude/skills/rohme-launchpad-interface curl -o .claude/skills/rohme-launchpad-interface/SKILL.md \ https://forum.rohme.capital/skills/rohme-launchpad-interface/SKILL.md
“Using the rohme-launchpad-interface skill, build me a launchpad explorer with a DAO list, per-DAO dashboards, and a swap panel, then deploy it.”
Build a dedicated, branded site for ONE launched DAO — its swap, staking, bonds, governance, treasury, and rage-quit surfaces. Your agent asks for your launchId and derives every address on-chain.
mkdir -p .claude/skills/rohme-dao-interface curl -o .claude/skills/rohme-dao-interface/SKILL.md \ https://forum.rohme.capital/skills/rohme-dao-interface/SKILL.md
“Using the rohme-dao-interface skill, build a standalone site for my DAO (launchId 3) with swap, staking, bonds, and governance, branded in our colors, then deploy it.”
Both skills instruct the agent to fetch — never transcribe — the canonical build inputs hosted alongside them:
The manifest and ABIs update when the protocol redeploys, so a fresh download always matches the live contracts. Contracts are also verified on Blockscout for independent cross-checking — see For Developers for the full address book.