# Agent Skill

Enable AI agents to interact with OmniHub directly from the terminal and access real onchain actions across the OmniHub ecosystem.

OmniHub Agent Skill is a modular knowledge and tooling package built for AI coding assistants, autonomous agents, and Web3 builders who want to create, mint, query, and manage digital collectibles faster.

***

### Installing the Skill

```
npx skills add omnihub-xyz/omnihub-skill -g -a claude-code -y
```

### **Manual Installation**

Clone this repository to your skills directory:

```
git clone https://github.com/omnihub-xyz/omnihub-skill.git ~/.claude/skills/omnihub-skill
```

Refer to your AI tool's documentation for skills directory configuration.

***

### Documentation

Full documentation including setup guides, operations reference, shell scripts, and configuration is maintained in the GitHub repository:

**GitHub:** <https://github.com/omnihub-xyz/omnihub-skill>

***

### What is OmniHub Agent Skill?

OmniHub Agent Skill gives AI agents a simple way to work with OmniHub.

Instead of manually navigating through every step, agents can use OmniHub Skill to perform core actions such as creating NFTs, minting collections, bridging tokens, swapping tokens, claiming testnet tokens from the faucet, and querying NFT collection data.

***

### Why use OmniHub Agent Skill?

OmniHub is a multichain launchpad built for speed, simplicity, and creativity. The Agent Skill extends this experience to AI agents, making OmniHub easier to access from developer environments, automation workflows, and agent-based products.

1. **Built for onchain AI agents:**\
   Designed for Claude Code, Cursor, Windsurf, and other coding agents that need real Web3 actions, not just static answers.
2. **Creator-first execution flows:**\
   Agents can create NFTs, mint collections, claim faucet tokens, bridge assets, swap tokens, and query NFT data through OmniHub workflows.
3. **Multichain by default:**\
   Use OmniHub across supported EVM networks, including mainnets and testnets, with one simple skill layer.
4. **Optimized for NFT creation:**\
   Built around OmniHub’s core experience: fast, no-code NFT creation, AI-powered collection building, and easy minting.
5. **Terminal-ready utility:**\
   Let agents interact with OmniHub directly from the terminal, making NFT and onchain actions faster for builders, creators, and ecosystem teams.
6. **Useful for campaigns and quests:**\
   Great for ecosystem activations where users need to create NFTs, mint collectibles, claim testnet tokens, or complete onchain tasks.
7. **Agent-friendly context:**\
   Structured to help AI agents understand OmniHub actions clearly and guide users through each step with less friction.

***

### How it works

Each request goes through these layers:

**\`SKILL.md\`**

Defines the canonical interfaces, chain resolution rules, AUTO\_CONFIRM behavior, and flow logic.

**Tool registry**

Maps commands like \`mint:prepare\`, \`collection:deploy\`, \`lifi:bridge\`, and \`omnihub:discover\` to typed runtime functions.

**Chain resolution**

Resolves user chain input through OmniHub’s \`/api/chains\` registry. Chain IDs, RPCs, factory addresses, faucet flags, and testnet flags are not hardcoded.

**Backend integration**

Handles OmniHub auth, faucet claims, collection sync, and collection metadata provisioning.

**Bridge and Swap by LI.FI integration**

Powers same-chain swaps and cross-chain bridges through a shared quote and execution flow.

***

### Collection creation limits

This skill currently supports only a basic collection creation flow.

**Supported:**

* mint price fixed to \`0\`
* one default public phase
* default royalty
* default supply
* automatic backend sync and metadata update

**Not supported:**

* custom mint price
* custom phases
* allowlists / Merkle roots
* advanced mint configuration

If those are needed, create the collection first and configure advanced settings manually afterward.

***

### Safety notes

* Real write flows sign real on-chain transactions.
* Product-choice prompts stay explicit where intended, even if auto-confirm is enabled.
* API responses are treated as untrusted input.
* Execution is EVM-only where applicable.
* OmniHub auth tokens are session-scoped and are not written to disk.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://omnihub-launchpad.gitbook.io/omnihub/api-and-agent-skill/agent-skill.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.