​

The Hyperdrome Agent

​

Not a DEX with AI — It’s an AI Agent with a DEX

Hyperdrome is built AI-first. The agent isn’t a chatbot bolted onto a trading interface — it is the interface. Every DeFi action (swapping, LPing, locking, voting, claiming) is accessible through natural language conversation or slash commands. The traditional UI exists as a fallback, but the agent is the primary way users interact with the protocol.

---

​

Architecture

The Hyperdrome agent uses an LLM routing layer to interpret user intent and map it to on-chain actions. When a user sends a message, the agent:

1. Parses intent — The LLM analyzes the message to determine what the user wants to do (swap, LP, vote, check portfolio, etc.)
2. Gathers context — Reads on-chain state (wallet balances, pool data, veNFT positions, pending rewards) and off-chain data (prices, TVL, APRs)
3. Generates response — Produces a natural language explanation with rich UI cards showing the proposed action, parameters, and expected outcomes
4. Executes on approval — When the user confirms, the agent constructs and submits the transaction(s)

The agent supports multiple LLM backends (Claude, GPT-4, Gemini) at zero cost to the user.

---

​

Chat-to-DeFi

Users interact with the full protocol through conversational prompts:

• “Swap 100 USDC to HYPE”
• “What are my pending rewards?”
• “Lock 1000 HDROME for 4 years”
• “Vote all my veHDROME on the HYPE/USDC pool”
• “What’s the best yield right now?”

The agent handles routing, slippage, gas estimation, and transaction construction behind the scenes.

---

​

Slash Commands

---

​

Rich UI Cards

The agent renders interactive cards for every action — not just text responses. Cards display:

• Swap cards — Token pair, amount, price impact, route path, slippage tolerance
• LP cards — Pool pair, TVL, APR, your share, deposit amounts
• Vote cards — Gauge list, current vote weights, projected fee/bribe revenue
• Portfolio cards — Token balances, LP positions, veNFT details, pending rewards
• Reward cards — Breakdown of claimable emissions, fees, and bribes with USD values

---

​

Executable Actions

---

​

Smart Suggestions

The agent proactively surfaces optimization opportunities based on wallet state:

---

​

LLM-Agnostic Architecture

The Hyperdrome agent is not locked into a single AI provider. The architecture separates intent parsing, context gathering, and execution into independent layers — so the LLM backend can be swapped without changing anything else. Currently supports Claude, GPT-4, and Gemini with intelligent routing. Open-source models (Llama, Mistral, Qwen) and Bring Your Own LLM are on the roadmap.

---

​

Multi-Language Support

The agent responds in the user’s language. Prompts in English, Chinese, Japanese, Korean, Spanish, French, Portuguese, Turkish, Russian, Arabic, and other languages are all supported natively through the underlying LLM capabilities. The UI labels and slash commands remain in English, but all conversational responses adapt to the user’s language.