ChainScope

Route questions to AI agents, query live data, and surface one-click transactions

ChainScope

Created At

ETHGlobal Lisbon 2026

Project Description

ChainScope is an intelligent orchestration platform designed to bridge the gap between complex data querying and actionable execution. When a user asks a question, the system dynamically routes the prompt to specialized AI agents. These agents seamlessly query live Graph data and analyze the retrieved information within a secure Python sandbox environment. These agents retrieve live blockchain information directly from The Graph using Subgraph MCP for discovery and querying of thousands of subgraphs, the Token API for unified balances and transfer history across chains Instead of merely returning a text response, ChainScope processes these insights to surface executable, one-click actions—transforming passive data retrieval into immediate, verifiable transactions.

How it's Made

ChainScope is built on a decoupled architecture optimized for speed, agentic orchestration, and deep visibility. The frontend is built in Next.js and deployed on Vercel for low-latency edge delivery. The backend API is powered by FastAPI and hosted on a Hetzner VPS, running behind Caddy as a reverse proxy for automated TLS termination and secure request routing. Complex, multi-agent state machines and sandboxed execution loops are constructed using LangChain and LangGraph. To monitor non-deterministic agent workflows, LangSmith is integrated across the backend, providing end-to-end trace observability, prompt latency tracking, and step-by-step state debugging.

background image mobile

Join the mailing list

Get the latest news and updates

ChainScope | ETHGlobal