Deep Dive
1. MCP Servers & AI Agent Skills (11 June 2026)
Overview: This update connects AI agents directly to on-chain data. Users can now ask questions in everyday language without needing to write complex code, making blockchain data accessible to a much wider audience.
The Graph released Managed Chain Provider (MCP) servers and new skills for AI agents like Claude and ChatGPT. These tools allow the agents to understand and query data from Subgraphs and Substreams. Instead of requiring developers to use GraphQL or SQL, the system interprets natural language requests, fetches the relevant live data, and returns it in a simple format.
What this means: This is bullish for GRT because it opens the protocol to massive new users—AI agents and non-technical people. It could lead to a surge in query volume as accessing data becomes as easy as having a conversation, driving more demand for The Graph's services and the GRT token used to pay for them.
(Source)
2. Horizon Subgraph Service Mainnet (Q1 2026)
Overview: This is a foundational upgrade to The Graph's core protocol. It transitions the network from a single-purpose indexing service to a flexible platform that can run multiple types of data services, all secured by GRT.
The Horizon upgrade, which launched on testnet in late 2025, introduces a modular blockchain architecture. Its mainnet rollout allows services like Subgraphs, Substreams (for real-time streaming), and the Token API to operate on a single, unified protocol. This creates a more efficient and scalable "data backbone" for Web3.
What this means: This is bullish for GRT because it future-proofs the network. By supporting specialized services for analytics, AI, and regulated finance, The Graph can capture more use cases and fee revenue. This diversification strengthens the utility of the GRT token across the entire ecosystem.
(Source)
3. Subgraph Dev Mode & New Features (30 October 2025)
Overview: This suite of tools drastically speeds up development for builders. Developers can now test changes to their data indexes instantly on their local machine, cutting out slow deployment steps.
The release includes four key features: Subgraph Dev Mode (local iteration without IPFS), Composition (reusing existing Subgraphs like building blocks), Aggregations (pre-computing daily/hourly trends), and Declarative eth_calls (parallel data fetching). Together, they reduce development time from hours to minutes.
What this means: This is bullish for GRT because it removes friction for developers. Faster, easier building leads to more Subgraphs being created and more applications relying on The Graph for data. This growth in the network's core product directly increases the value and usage of the GRT token.
()
Conclusion
The Graph's development trajectory is sharply focused on becoming a modular, AI-native data layer, with recent updates lowering barriers for both developers and end-users. How will the integration of natural language queries reshape the competitive landscape for blockchain data providers?