AI Revamp (#987)

* working draft

* Add icon to “AI Suite” section

* updates

* Edits

* Updating

* More Updates

* Edits

* Adding disclaimer

* Minor edits

* touchups

---------

Co-authored-by: benface <[email protected]>

Author: idalithb

nginx.conf

@@ -161,6 +161,13 @@ http {
         rewrite ^/docs/([a-zA-Z][a-zA-Z])/tokenomics/$                                  $scheme://$http_host/docs/$1/resources/tokenomics/                                  permanent;
         rewrite ^/docs/([a-zA-Z][a-zA-Z])/token-api/$                                   $scheme://$http_host/docs/$1/token-api/quick-start/                                 permanent;
         rewrite ^/docs/([a-zA-Z][a-zA-Z])/token-api/evm/get-transfers-evm-by-address/$  $scheme://$http_host/docs/$1/token-api/evm/get-transfers-evm/                       permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/subgraphs/mcp/claude/$                        $scheme://$http_host/docs/$1/ai-suite/subgraph-mcp/                                 permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/subgraphs/mcp/cline/$                         $scheme://$http_host/docs/$1/ai-suite/subgraph-mcp/                                 permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/subgraphs/mcp/cursor/$                         $scheme://$http_host/docs/$1/ai-suite/subgraph-mcp/                                permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/subgraphs/mcp/cline/$                         $scheme://$http_host/docs/$1/ai-suite/subgraph-mcp/                                 permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/token-api/mcp/cline/$                         $scheme://$http_host/docs/$1/ai-suite/token-api-mcp/                                permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/token-api/mcp/cline/$                         $scheme://$http_host/docs/$1/ai-suite/token-api-mcp/                                permanent;
+        rewrite ^/docs/([a-zA-Z][a-zA-Z])/token-api/mcp/cline/$                         $scheme://$http_host/docs/$1/ai-suite/token-api-mcp/                                permanent;
         # Temporary redirects (302)
         rewrite ^/docs/en/querying/graph-client/$                                       $scheme://$http_host/docs/en/subgraphs/querying/graph-client/README/                redirect;
         rewrite ^/docs/en/developing/graph-ts/$                                         $scheme://$http_host/docs/en/subgraphs/developing/creating/graph-ts/README/         redirect;

website/next.config.js

@@ -70,6 +70,13 @@ const withNextra = nextra({
         type: 'children',
         title: t('global.navigation.tokenApi'),
       },
+      '---4': {
+        type: 'separator',
+      },
+      'ai-suite': {
+        type: 'children',
+        title: t('global.navigation.ai-suite'),
+      },
       '---5': {
         type: 'separator',
       },

website/src/layout/Layout.tsx

@@ -43,6 +43,7 @@ import {
   SocialGitHub,
   SocialTelegram,
   SocialX,
+  Sparkle,
   Stack,
   Subgraph,
   Substreams,
@@ -198,6 +199,9 @@ export default function Layout({ pageOpts, children }: NextraThemeLayoutProps<Fr
         if (routeWithoutLocale === '/token-api' || routeWithoutLocale.startsWith('/token-api/')) {
           return <APIToken alt="" />
         }
+        if (routeWithoutLocale === '/ai-suite' || routeWithoutLocale.startsWith('/ai-suite/')) {
+          return <Sparkle alt="" variant={selected ? 'fill' : 'regular'} />
+        }
         if (routeWithoutLocale === '/indexing' || routeWithoutLocale.startsWith('/indexing/')) {
           return <RoleIndexer alt="" />
         }

website/src/pages/en/ai-suite/_meta-titles.json

@@ -0,0 +1,6 @@
+{
+  "graph-assistant": "Graph Assistant",
+  "ai-introduction": "The Graph's AI",
+  "subgraph-mcp": "Subgraph MCP",
+  "token-api-mcp": "Token API MCP"
+}

website/src/pages/en/ai-suite/_meta.js

@@ -0,0 +1,8 @@
+import titles from './_meta-titles.json'
+
+export default {
+  'ai-introduction': titles['ai-introduction'],
+  'graph-assistant': titles['graph-assistant'],
+  'subgraph-mcp': titles['subgraph-mcp'],
+  'token-api-mcp': titles['token-api-mcp'],
+}

website/src/pages/en/ai-suite/ai-introduction.mdx

@@ -0,0 +1,60 @@
+---
+title: The Graph's AI
+---
+
+Build with speed and scale faster with The Graph's AI.
+
+## Using AI on The Graph
+
+Instead of relying on static datasets or centralized APIs, you can now access live blockchain data through our **agentic** app, [Graph Assistant](/ai-suite/graph-assistant/introduction/), and use our AI-native tooling, [Subgraph MCP](/ai-suite/subgraph-mcp/introduction/), and [Token API MCP](/ai-suite/token-api-mcp/introduction/).
+
+### Why Use Onchain Data with AI?
+
+Using onchain data with AI unlocks powerful new ways to interact with and understand blockchain ecosystems.
+
+- **AI for Non-Technical End Users**: AI can turn complex onchain data into accessible insights. Assistants powered by AI enable you to explore and analyze blockchain data without coding.
+- **AI for Developers**: You can use AI to interact directly with The Graph's data through your agents or build AI-powered applications on top of it. This streamlines development and opens up more intuitive, dynamic use cases.
+
+## AI for Non-Technical/End Users
+
+[Graph Assistant](/ai-suite/graph-assistant/introduction/) is a powerful **agentic** application that empowers users to easily interact with blockchain data. Ask questions in simple, natural language and get rich, AI powered insights. Both technical developers and non-technical product teams can move faster, make smarter decisions, and gain a clearer understanding than ever before.
+
+### Benefits of Using Graph Assistant
+
+- **Natural language Input**: You don't need to write complex queries or sift through dashboards. Simply ask your question in plain English and get clear, structured results.
+- **Multi-source support**: Works across multiple Subgraphs and Token API. The Assistant automatically determines which data source to use, so you don't have to switch tools or manage endpoints.
+- **Schema-free access**: You don't need to understand the schema. You can refer to tokens, protocols, dates, or addresses using natural language. The Assistant takes care of the rest.
+
+### Conversational Querying with Graph Assistant
+
+**Example queries:**
+
+- "List the top-5 swaps by USD value in the last 24 hours."
+- "Give me the daily volume on Uniswap."
+- "List the 20 most recently registered .eth names."
+- "What's the best liqudity pool to put my ETH into?"
+- "What is the total delegated stake in The Graph network?"
+
+[Graph Assistant](/ai-suite/graph-assistant/introduction/) will eventually be a one-stop shop for interacting with The Graph. Right now, it's an agentic AI app that uses state-of-the-art AI techniques, LLMs, and MCPs to understand and answer your questions about [Subgraphs](/subgraphs/developing/subgraphs/) and [Token API](/token-api/quick-start/) data.
+
+## AI for Developers
+
+### Enable Natural Language Access to Onchain Data
+
+[Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) servers connect to Claude, Cline, and Cursor. They enable models to understand, query, and interact with structured onchain data using natural language. MCPs remove the need to write low-level queries or interact with APIs directly.
+
+### Subgraph MCP
+
+The [Subgraph MCP](/ai-suite/subgraph-mcp/introduction/) server connects models to Subgraphs on The Graph Network. It allows language models to explore Subgraph schemas, execute GraphQL queries, find relevant Subgraphs by keyword or contract, and surface usage metrics using natural language.
+
+### Token API MCP
+
+The [Token API MCP](/ai-suite/token-api-mcp/introduction/) MCP server enables models to access onchain token data. It supports querying token metadata, wallet balances, transfer history, and top token holders through simple prompts.
+
+#### Benefits of Using Subgraph and Token API MCPs
+
+- Query Subgraphs and token data using natural language
+- Access GraphQL schemas and token metadata without manual queries
+- Find Subgraphs by keyword or contract and check 30-day usage
+- Retrieve real-time data from The Graph and onchain token sources
+- Get wallet balances, transfer history, and top token holders

website/src/pages/en/ai-suite/graph-assistant/_meta.js

@@ -0,0 +1,6 @@
+export default {
+  'quick-start': '',
+  introduction: '',
+  'how-to-guide': '',
+  faq: '',
+}

website/src/pages/en/ai-suite/graph-assistant/faq.mdx

@@ -0,0 +1,26 @@
+---
+title: Graph Assistant FAQ
+sidebarTitle: 'FAQ'
+---
+
+This page summarizes some of the most common questions for developers using Graph Assistant.
+
+1. What is The Graph Assistant?
+
+The Graph Assistant is an AI-powered chat interface that lets you query blockchain data in plain English. Under the hood, it translates your questions into GraphQL calls against Subgraphs and the Token API, then returns results, no coding required.
+
+2. What is an agentic application?
+
+An agentic application combines AI-driven reasoning with external data sources or APIs to take autonomous actions on your behalf. In this case, Graph Assistant acts as an “agent” that interprets your natural-language prompts, builds and dispatches GraphQL queries, and formats the results so that you don't need to write or manage queries yourself.
+
+3. Which networks and data does Graph Assistant support?
+
+The Assistant has access to Subgraphs across multiple chains. If a Subgraph is published to The Graph Network and is being indexed, the Assistant can automatically route queries to it. The Assistant also has access to the Token API on Ethereum Mainnet, BSC, Unichain, Arbitrum-One, Optimism, and Polygon.
+
+4. Is there a usage limit or quota?
+
+Yes, 10 questions per hour.
+
+5. What if the Assistant can't interpret my question?
+
+When the Assistant fails to parse your intent (for example, ambiguous token name, missing context, or an unsupported request), it will prompt you for clarification. For instance, it might ask, “Did you mean XYZ on Ethereum or XYZ on Arbitrum?” or “Please specify a date range for this query.” If you continue to see parsing errors, check that you're using correct token symbols or specifying networks and time frames clearly.

website/src/pages/en/ai-suite/graph-assistant/how-to-guide.mdx

@@ -0,0 +1,56 @@
+---
+title: How To Use The Graph Assistant
+sidebarTitle: 'How-to Guide'
+---
+
+## Overview
+
+You type a natural language question into the chat interface, such as:
+
+```yaml
+'On Uniswap, return the top-10 pools by total value locked in USD.'
+```
+
+> [!IMPORTANT] All users can process 10 free queries per user per hour.
+
+> [!NOTE] Disclaimer: By using Graph Assistant, you acknowledge that your requests may be logged and used to improve the performance of the underlying agents. Do not share sensitive, personal, or private information.
+
+### Behind the Scenes
+
+The Assistant:
+
+1. Interprets your intent
+2. Maps it to the correct Subgraphs or Token API endpoint
+3. Executes optimized queries
+4. Returns clean, structured results
+
+You don't need to write queries or understand a Subgraph's schema, you simply ask.
+
+### Example Use Cases
+
+**Protocol Analytics**
+
+- “Compare daily active users on Aave vs. Compound in May 2025.”
+- “Show me the total value locked (TVL) for XYZDAO over the last quarter.”
+
+**Token Insights**
+
+- “What are the top 10 tokens by market cap on Ethereum right now?”
+- “How many XYZ did I receive between May 1 and May 15?”
+
+**Historical Data Retrieval**
+
+- “Fetch the token price history of XYZ for the last six months.”
+- “What was the average gas price on Ethereum during April?”
+
+**Portfolio Monitoring**
+
+- “List my wallet’s current token balances and their USD equivalents.”
+- “Calculate profit/loss on XYZ since xyz date.”
+
+**Cross-Chain Comparison**
+
+- “Show me the 7-day trading volume of XYZ on Ethereum vs. Polygon.”
+- “How many XYZ swaps occurred on Arbitrum yesterday?”
+
+To get started follow the [quickstart](/ai-suite/graph-assistant/quick-start/) and explore its capabilities.

website/src/pages/en/ai-suite/graph-assistant/introduction.mdx

@@ -0,0 +1,31 @@
+---
+title: Introduction
+---
+
+> [!IMPORTANT] Graph Assistant is still in **beta**. Expect breaking changes, rate limits, and potential downtime.
+
+[Graph Assistant](https://thegraph.com/graph-assistant) (Beta) gives instant, code-free access to blockchain data, transforming natural language into precise, AI-powered queries without the need to code.
+
+## Overview
+
+[Graph Assistant](https://thegraph.com/graph-assistant) is an advanced, [agentic](/en/ai-suite/graph-assistant/faq/) easy-to-use chat interface for querying blockchain data using natural language. Unlike a basic chatbot, Graph Assistant can understand your intent, build complex queries, and execute them automatically. This gives you rich, AI-powered insights from onchain data without writing any code.
+
+- **Automatically** interprets user intent, generates precise queries, executes them via The Graph's Subgraphs and Token API, and presents structured, AI-powered answers without requiring any code.
+- **Removes** the technical complexity found in traditional dashboards or chatbots. It provides composable insights directly from indexed Subgraphs and Token datasets.
+
+## Core Features
+
+| Feature | Description |
+| --- | --- |
+| **Friendly Interface** | Conversational access to blockchain data via natural language. You simply ask questions, the Assistant fetches the answer, and renders results. |
+| **Backed by Subgraphs** | Uses subgraphs to index protocol-specific data like transactions, events, and protocol metrics. |
+| **Backed by Token API** | Uses Token API for token-level information like balances, transfers, and metadata. |
+| **No Setup Required** | No need to configure data sources or write queries. |
+| **Automatic Query Building** | The Assistant performs complex queries such as fetching historical token volumes, protocol analytics, or cross-chain data and handles filtering, pagination, and joins across Subgraphs automatically. |
+
+## Help Shape What's Next
+
+This is in **beta**, and community input is essential. If you have ideas, questions, or run into anything unexpected, we'd love to hear from you:
+
+- Join the community on [Discord](https://discord.com/invite/graphprotocol): Provide feedback and stay in the loop with other builders. You'll be able to join the `#graph-assistant` channel to share your experience directly.
+- Interested in going deeper? Reach out to become a dedicated beta tester so your specific use case can be better supported.

website/src/pages/en/ai-suite/graph-assistant/quick-start.mdx

@@ -0,0 +1,44 @@
+---
+title: Quick Start
+---
+
+> [!IMPORTANT] Graph Assistant is still in **beta**. Expect breaking changes, rate limits, and potential downtime.
+
+This guide provides step-by-step instructions for getting started with Graph Assistant quickly.
+
+## Step-by Step
+
+### Step 1. Create an Account
+
+Go to: [Graph Assistant](https://thegraph.com/graph-assistant)
+
+- Click Sign Up (or Log In if you already have an account).
+- Use your Google, GitHub, X, email or wallet credentials to log in.
+- No additional setup is required.
+
+### Step 2. Open the Assistant Chat
+
+After logging in:
+
+- You'll see Graph Assistant's interface.
+- At the bottom of the page, you will find a chat input bar. This is where you will interact with the Assistant.
+
+### Step 3: Ask Your First Question
+
+> [!IMPORTANT] All users can process 10 free queries per user per hour.
+
+In the input bar, type a natural-language query (everyday language).
+
+Example:
+
+```yaml
+Show me the 24-hour trading volume of XYZ token on Ethereum
+```
+
+> [!NOTE] Disclaimer: By using Graph Assistant, you acknowledge that your requests may be logged and used to improve the performance of the AI interface. Do not share sensitive, personal, or private information.
+
+The Assistant will:
+
+- Digest your question
+- Fetch onchain data using Subgraphs or the Token API
+- Return a clean, formatted response with the data you need

website/src/pages/en/subgraphs/mcp/_meta.js renamed to website/src/pages/en/ai-suite/subgraph-mcp/_meta.js

@@ -1,4 +1,5 @@
 export default {
+  introduction: '',
   claude: '',
   cline: '',
   cursor: '',

website/src/pages/en/subgraphs/mcp/claude.mdx renamed to website/src/pages/en/ai-suite/subgraph-mcp/claude.mdx

@@ -2,17 +2,7 @@
 title: Claude Desktop
 ---
 
-The Subgraph [Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) server enables Claude to interact directly with Subgraphs on The Graph Network. This integration allows you to find relevant Subgraphs for specific keywords and contracts, explore Subgraph schemas, and execute GraphQL queries—all through natural language conversations with Claude.
-
-## What You Can Do
-
-The Subgraph MCP integration enables you to:
-
-- Access the GraphQL schema for any Subgraph on The Graph Network
-- Execute GraphQL queries against any Subgraph deployment
-- Find top Subgraph deployments for a given keyword or contract address
-- Get 30-day query volume for Subgraph deployments
-- Ask natural language questions about Subgraph data without writing GraphQL queries manually
+This guide walks you through the integration of Subgraph [Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) to access Subgraph data via natural language conversations with Claude.
 
 ## Prerequisites
 

website/src/pages/en/subgraphs/mcp/cline.mdx renamed to website/src/pages/en/ai-suite/subgraph-mcp/cline.mdx

@@ -2,7 +2,7 @@
 title: Cline
 ---
 
-The Subgraph [Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) server enables Cline to interact directly with Subgraphs on The Graph Network. This integration allows you to explore Subgraph schemas, execute GraphQL queries, and find relevant Subgraphs for specific contracts—all through natural language conversations with Cline.
+This guide walks you through the integration of Subgraph [Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) to access Subgraph data through natural language conversations with Cline.
 
 ## Prerequisites
 

website/src/pages/en/subgraphs/mcp/cursor.mdx renamed to website/src/pages/en/ai-suite/subgraph-mcp/cursor.mdx

@@ -2,7 +2,7 @@
 title: Cursor
 ---
 
-The Subgraph [Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) server enables Cursor to interact directly with Subgraphs on The Graph Network. This integration allows you to explore Subgraph schemas, execute GraphQL queries, and find relevant Subgraphs for specific contracts—all through natural language conversations with Cursor.
+This guide walks you through the integration of Subgraph [Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) to access Subgraph data through natural language conversations with Cline.
 
 ## Prerequisites
 

website/src/pages/en/ai-suite/subgraph-mcp/introduction.mdx

@@ -0,0 +1,22 @@
+---
+title: Subgraph MCP Introduction
+sidebarTitle: 'Introduction'
+---
+
+Subgraph MCP server is an open-source implementation of [Anthropic's Model Context Protocol](https://modelcontextprotocol.io/introduction).
+
+It presents The Graph's Subgraph data through a set of MCP tools that any MCP-compatible client can call. With these tools, clients can search for relevant Subgraphs, inspect GraphQL schemas, and run queries against specific deployments on The Graph Network all within a single, standardized interface.
+
+The server itself does not hold a language model. Instead, it translates incoming MCP requests into Subgraph queries and returns the structured results, allowing the client's LLM to transform a natural-language prompt into actionable blockchain data.
+
+Think of it as a USB-C hub: it standardizes the plug-and-play connection between AI agents and The Graph's Subgraph ecosystem.
+
+## What You Can Do
+
+- Access GraphQL schemas for any Subgraph on The Graph Network
+- Run GraphQL queries on any Subgraph deployment
+- Discover top Subgraph deployments by keyword or contract address
+- Retrieve 30-day query volumes for Subgraph deployments
+- Ask questions about Subgraph data without writing GraphQL manually
+
+The Subgraph MCP server allows smooth integration with [Claude](/ai-suite/subgraph-mcp/claude/), [Cline](/ai-suite/subgraph-mcp/cline/), or [Cursor](/ai-suite/subgraph-mcp/cursor/), making blockchain data queries with The Graph Network a conversational experience.

website/src/pages/en/token-api/mcp/_meta.js renamed to website/src/pages/en/ai-suite/token-api-mcp/_meta.js

@@ -1,4 +1,5 @@
 export default {
+  introduction: '',
   claude: '',
   cline: '',
   cursor: '',