explaingit

vexorterminal0111/vexor-terminal

0TypeScriptAudience · developerComplexity · 4/5ActiveLicenseSetup · hard

TLDR

Chat product wrapped around a $VT ERC-20 on Base, where a Llama 3.3 70B orchestrator on Groq routes user messages to nine themed sub-agents and token holders stake to vote.

Mindmap

mindmap
  root((vexor-terminal))
    Inputs
      User chat
      Wallet signatures
      Staking actions
    Outputs
      AI replies
      Token rewards
      On chain votes
    Use Cases
      Run a tokenized chat dApp
      Stake $VT for revenue share
      Vote on protocol changes
    Tech Stack
      Next.js
      Solidity
      Foundry
      Cloudflare Workers
      Groq

Things people build with this

USE CASE 1

Spin up a token-gated AI chat where a primary LLM routes to themed sub-agents

USE CASE 2

Study a Next.js 16 site exported static and served from a single Cloudflare Worker with a /api/chat proxy to Groq

USE CASE 3

Reuse the OpenZeppelin-based staking and voting contracts for a Base ERC-20

USE CASE 4

Test the lock-period revenue share flow against Base Sepolia before mainnet

Tech stack

TypeScriptNext.jsSolidityFoundryCloudflare WorkersGroqwagmi

Getting it running

Difficulty · hard Time to first run · 1day+

Standing it up needs pnpm, a Groq API key, a Cloudflare Worker for the chat proxy, and Foundry plus a Base Sepolia wallet to exercise staking and voting.

MIT, use, copy, modify, and ship for any purpose as long as the copyright notice stays.

In plain English

Vexor Terminal is a website plus a set of blockchain contracts that together form a chat product wrapped around a cryptocurrency token. On the surface a user opens vexorterminal.com, optionally connects a crypto wallet, and chats with an AI. The README presents the AI side as an orchestrator: a single language model receives the user's message and forwards it to one of nine themed sub agents, each named (Cipher, Atlas, Quill and so on) and each running its own model. Under the hood the chat replies are generated by Llama 3.3 70B running on a third party inference service called Groq. The token side runs on Base, an Ethereum compatible blockchain operated by Coinbase. There is a live ERC 20 token called $VT on Base mainnet with a fixed supply of 100 billion units, and a separate test version on Base Sepolia for trying the staking and voting features without spending real money. Holders can lock their tokens for fixed periods of flex, 30, 90, or 180 days to earn a share of revenue and to gain weighted voting power over future protocol changes. The voting layer uses well known open source contracts from OpenZeppelin. The website is built with Next.js 16, exported as static files, and served from a single Cloudflare Worker that also hosts a small proxy at /api/chat which forwards messages to Groq. Wallet interactions use wagmi, viem, and RainbowKit. Smart contracts are written in Solidity 0.8.26 and tested with Foundry, with the README reporting 9 of 9 tests passing. Quick start instructions cover cloning the repo, installing dependencies with pnpm, copying an example environment file, and running the development server on port 3000. The project is released under the MIT license.

Copy-paste prompts

Prompt 1
Clone vexor-terminal, set up the .env file, and run the Next.js dev server on port 3000 against Base Sepolia
Prompt 2
Deploy the $VT staking contracts to Base Sepolia with Foundry and walk me through locking tokens for the 90-day tier
Prompt 3
Swap the Groq Llama 3.3 70B backend in vexor-terminal for a local Ollama model behind the same /api/chat proxy
Prompt 4
Add a tenth sub-agent to the vexor-terminal orchestrator that specializes in on-chain Solidity questions
Open on GitHub → Explain another repo

Generated 2026-05-22 · Model: sonnet-4-6 · Verify against the repo before relying on details.