# How to integrate Zenrows MCP with Vercel AI SDK v6 ```json { "title": "How to integrate Zenrows MCP with Vercel AI SDK v6", "toolkit": "Zenrows", "toolkit_slug": "zenrows", "framework": "Vercel AI SDK", "framework_slug": "ai-sdk", "url": "https://composio.dev/toolkits/zenrows/framework/ai-sdk", "markdown_url": "https://composio.dev/toolkits/zenrows/framework/ai-sdk.md", "updated_at": "2026-05-06T08:34:34.376Z" } ``` ## Introduction This guide walks you through connecting Zenrows to Vercel AI SDK v6 using the Composio tool router. By the end, you'll have a working Zenrows agent that can download a pdf of this news article, extract plain text from the given webpage, get latest property data from zillow through natural language commands. This guide will help you understand how to give your Vercel AI SDK agent real control over a Zenrows account through Composio's Zenrows MCP server. Before we dive in, let's take a quick look at the key ideas and tools involved. ## Also integrate Zenrows with - [OpenAI Agents SDK](https://composio.dev/toolkits/zenrows/framework/open-ai-agents-sdk) - [Claude Agent SDK](https://composio.dev/toolkits/zenrows/framework/claude-agents-sdk) - [Claude Code](https://composio.dev/toolkits/zenrows/framework/claude-code) - [Claude Cowork](https://composio.dev/toolkits/zenrows/framework/claude-cowork) - [Codex](https://composio.dev/toolkits/zenrows/framework/codex) - [OpenClaw](https://composio.dev/toolkits/zenrows/framework/openclaw) - [Hermes](https://composio.dev/toolkits/zenrows/framework/hermes-agent) - [CLI](https://composio.dev/toolkits/zenrows/framework/cli) - [Google ADK](https://composio.dev/toolkits/zenrows/framework/google-adk) - [LangChain](https://composio.dev/toolkits/zenrows/framework/langchain) - [Mastra AI](https://composio.dev/toolkits/zenrows/framework/mastra-ai) - [LlamaIndex](https://composio.dev/toolkits/zenrows/framework/llama-index) - [CrewAI](https://composio.dev/toolkits/zenrows/framework/crew-ai) ## TL;DR Here's what you'll learn: - How to set up and configure a Vercel AI SDK agent with Zenrows integration - Using Composio's Tool Router to dynamically load and access Zenrows tools - Creating an MCP client connection using HTTP transport - Building an interactive CLI chat interface with conversation history management - Handling tool calls and results within the Vercel AI SDK framework ## What is Vercel AI SDK? The Vercel AI SDK is a TypeScript library for building AI-powered applications. It provides tools for creating agents that can use external services and maintain conversation state. Key features include: - streamText: Core function for streaming responses with real-time tool support - MCP Client: Built-in support for Model Context Protocol via @ai-sdk/mcp - Step Counting: Control multi-step tool execution with stopWhen: stepCountIs() - OpenAI Provider: Native integration with OpenAI models ## What is the Zenrows MCP server, and what's possible with it? The Zenrows MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Zenrows account. It provides structured and secure access to advanced web scraping capabilities, so your agent can extract structured data, bypass CAPTCHAs, convert pages to PDF, and monitor your API usage on your behalf. - Intelligent web data extraction: Direct your agent to scrape and extract plain text or structured data from dynamic websites, including specialized real estate property data from platforms like Zillow or Idealista. - PDF and content generation: Ask your agent to convert any web page into a PDF or retrieve clean, formatted plain text for archiving, documentation, or offline reading. - Seamless CAPTCHA and block bypassing: Enable your agent to gather data from sites protected by CAPTCHAs or anti-bot systems without manual intervention. - Real-time API usage monitoring: Have the agent check your account’s current API usage, concurrency status, and limits to help manage credits and avoid interruptions. - Session and compression management: Instruct your agent to maintain consistent scraping sessions, handle compression to optimize bandwidth, and retrieve detailed response headers for debugging and performance optimization. ## Supported Tools | Tool slug | Name | Description | |---|---|---| | `ZENROWS_GET_API_USAGE` | Get ZenRows API Usage Statistics | This tool retrieves the current api usage statistics and limits for your zenrows account. it is an independent action that requires no additional parameters besides authentication. it is useful for monitoring api usage and ensuring that the account has sufficient remaining credits. | | `ZENROWS_GET_CONCURRENCY_STATUS` | Get Concurrency Status | This tool retrieves the current concurrency status of your zenrows api usage. it reports the maximum number of concurrent requests allowed by your plan and the number of available concurrent request slots. it is useful for monitoring api usage, implementing rate limiting, debugging request issues, and capacity planning. | | `ZENROWS_GET_CONCURRENCY_STATUS_DETAILED` | Get Detailed Concurrency Status | This tool provides detailed information about the current concurrency status and limits of your zenrows account by making a request to the api and analyzing the response headers. it is essential for monitoring real-time api usage, managing concurrent requests, and ensuring optimal performance within plan limits. | | `ZENROWS_GET_ORIGINAL_STATUS` | Get Original Status Code | This tool retrieves the original http status code returned by the target website, which is useful for debugging purposes. it returns the original status code in the response headers under 'x-zenrows-original-status'. it can also provide the full response body and error messages, helping with debugging scraping issues, verifying website responses, monitoring availability, and understanding website behavior. | | `ZENROWS_GET_PDF_FROM_URL` | Get PDF from URL | This tool generates a pdf version of the scraped content from a given url. it requires javascript rendering to be enabled and sets the response type to pdf, making it ideal for archiving web pages, creating documentation, generating reports, or saving articles for offline reading. | | `ZENROWS_GET_PLAINTEXT` | Get Plaintext Response | This tool extracts plain text content from a webpage using the zenrows api. by using the response type=plaintext parameter, it strips html tags and formats the content into clean, plain text. it's useful for extracting readable content for nlp, summarization, or archiving purposes. | | `ZENROWS_GET_REAL_ESTATE_DATA` | Get Real Estate Property Data | A specialized tool for extracting structured data from real estate platforms like zillow and idealista. it leverages zenrows' real estate api to fetch comprehensive property information, including property details, location information, features, seller details, and more, in a structured format. | | `ZENROWS_GET_RESPONSE_COMPRESSION` | Get Response with Compression | A tool to fetch content from a url using the zenrows api with compression enabled to optimize bandwidth usage and improve performance. it supports gzip, deflate, and br compression encodings, handles decompression automatically, and provides compression statistics along with the decompressed content. | | `ZENROWS_GET_RESPONSE_HEADERS` | Get response headers | A tool to retrieve and parse response headers from zenrows api requests. it provides critical metadata such as concurrency limits, available request slots, request cost, unique request id, and final url after redirects, which is essential for monitoring usage, debugging, and optimizing requests. | | `ZENROWS_GET_SESSION_ID` | Get Session ID | This tool implements zenrows' session management functionality to maintain the same ip address across multiple requests for up to 10 minutes. it supports parameters like url, session id, and premium proxy, and is useful for maintaining consistent scraping sessions, simulating real user behavior, and avoiding detection by anti-bot systems. | | `ZENROWS_GET_WALMART_PRODUCT` | Get Walmart Product Details | This tool allows users to extract detailed product information from walmart using zenrows' specialized e-commerce scraping api. it provides structured data for walmart products including product details, pricing, availability, and more. | | `ZENROWS_SCRAPE_URL` | Scrape url | Scrape and extract data from a specified url. this action allows you to collect and process web data effortlessly using the zenrows api. | | `ZENROWS_SCRAPE_URL_AUTOPARSE` | Scrape url autoparse | The zenrows scrape url autoparse tool automatically parses and extracts structured data from any given url using intelligent parsing capabilities. it eliminates the need for manual css selectors by auto-identifying relevant content on web pages, returning data such as titles, main content, meta descriptions, images, links, prices, and contact information in a structured json format. | | `ZENROWS_SCRAPE_URL_HTML` | Scrape URL HTML | This tool extracts raw html data from a given url using zenrows' universal scraper api. it focuses on retrieving the pure html content of the webpage without automatic parsing or data extraction. it supports parameters such as js render for enabling javascript rendering, custom headers for custom http headers, premium proxy for using premium proxies, and session id for maintaining the same ip across multiple requests. | | `ZENROWS_SCRAPE_WITH_CSS_SELECTORS` | Scrape URL with CSS Selectors | This tool allows users to scrape specific elements from a webpage using css selectors. it is particularly useful for targeted data extraction rather than retrieving the entire page content. the endpoint takes a url and a json object containing css selectors for parsing elements such as titles, links, images, and prices, and includes optional parameters like using premium proxies, specifying response wait times, and custom headers among others. | | `ZENROWS_SCREENSHOT_URL` | Screenshot URL | A tool to capture screenshots of web pages using zenrows api. this tool allows you to take screenshots of entire web pages or specific elements, with customizable options for format and quality. | ## Supported Triggers None listed. ## Creating MCP Server - Stand-alone vs Composio SDK The Zenrows MCP server is an implementation of the Model Context Protocol that connects your AI agent to Zenrows. It provides structured and secure access so your agent can perform Zenrows operations on your behalf through a secure, permission-based interface. With Composio's managed implementation, you don't have to create your own developer app. For production, if you're building an end product, we recommend using your own credentials. The managed server helps you prototype fast and go from 0-1 faster. ## Step-by-step Guide ### 1. Prerequisites Before you begin, make sure you have: - Node.js and npm installed - A Composio account with API key - An OpenAI API key ### 1. Getting API Keys for OpenAI and Composio OpenAI API Key - Go to the [OpenAI dashboard](https://platform.openai.com/settings/organization/api-keys) and create an API key. You'll need credits to use the models, or you can connect to another model provider. - Keep the API key safe. Composio API Key - Log in to the [Composio dashboard](https://dashboard.composio.dev?utm_source=toolkits&utm_medium=framework_docs). - Navigate to your API settings and generate a new API key. - Store this key securely as you'll need it for authentication. ### 2. Install required dependencies First, install the necessary packages for your project. What you're installing: - @ai-sdk/openai: Vercel AI SDK's OpenAI provider - @ai-sdk/mcp: MCP client for Vercel AI SDK - @composio/core: Composio SDK for tool integration - ai: Core Vercel AI SDK - dotenv: Environment variable management ```bash npm install @ai-sdk/openai @ai-sdk/mcp @composio/core ai dotenv ``` ### 3. Set up environment variables Create a .env file in your project root. What's needed: - OPENAI_API_KEY: Your OpenAI API key for GPT model access - COMPOSIO_API_KEY: Your Composio API key for tool access - COMPOSIO_USER_ID: A unique identifier for the user session ```bash OPENAI_API_KEY=your_openai_api_key_here COMPOSIO_API_KEY=your_composio_api_key_here COMPOSIO_USER_ID=your_user_id_here ``` ### 4. Import required modules and validate environment What's happening: - We're importing all necessary libraries including Vercel AI SDK's OpenAI provider and Composio - The dotenv/config import automatically loads environment variables - The MCP client import enables connection to Composio's tool server ```typescript import "dotenv/config"; import { openai } from "@ai-sdk/openai"; import { Composio } from "@composio/core"; import * as readline from "readline"; import { streamText, type ModelMessage, stepCountIs } from "ai"; import { createMCPClient } from "@ai-sdk/mcp"; const composioAPIKey = process.env.COMPOSIO_API_KEY; const composioUserID = process.env.COMPOSIO_USER_ID; if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set"); if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set"); if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set"); const composio = new Composio({ apiKey: composioAPIKey, }); ``` ### 5. Create Tool Router session and initialize MCP client What's happening: - We're creating a Tool Router session that gives your agent access to Zenrows tools - The create method takes the user ID and specifies which toolkits should be available - The returned mcp object contains the URL and authentication headers needed to connect to the MCP server - This session provides access to all Zenrows-related tools through the MCP protocol ```typescript async function main() { // Create a tool router session for the user const session = await composio.create(composioUserID!, { toolkits: ["zenrows"], }); const mcpUrl = session.mcp.url; ``` ### 6. Connect to MCP server and retrieve tools What's happening: - We're creating an MCP client that connects to our Composio Tool Router session via HTTP - The mcp.url provides the endpoint, and mcp.headers contains authentication credentials - The type: "http" is important - Composio requires HTTP transport - tools() retrieves all available Zenrows tools that the agent can use ```typescript const mcpClient = await createMCPClient({ transport: { type: "http", url: mcpUrl, headers: session.mcp.headers, // Authentication headers for the Composio MCP server }, }); const tools = await mcpClient.tools(); ``` ### 7. Initialize conversation and CLI interface What's happening: - We initialize an empty messages array to maintain conversation history - A readline interface is created to accept user input from the command line - Instructions are displayed to guide the user on how to interact with the agent ```typescript let messages: ModelMessage[] = []; console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n"); console.log( "Ask any questions related to zenrows, like summarize my last 5 emails, send an email, etc... :)))\n", ); const rl = readline.createInterface({ input: process.stdin, output: process.stdout, prompt: "> ", }); rl.prompt(); ``` ### 8. Handle user input and stream responses with real-time tool feedback What's happening: - We use streamText instead of generateText to stream responses in real-time - toolChoice: "auto" allows the model to decide when to use Zenrows tools - stopWhen: stepCountIs(10) allows up to 10 steps for complex multi-tool operations - onStepFinish callback displays which tools are being used in real-time - We iterate through the text stream to create a typewriter effect as the agent responds - The complete response is added to conversation history to maintain context - Errors are caught and displayed with helpful retry suggestions ```typescript rl.on("line", async (userInput: string) => { const trimmedInput = userInput.trim(); if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) { console.log("\nGoodbye!"); rl.close(); process.exit(0); } if (!trimmedInput) { rl.prompt(); return; } messages.push({ role: "user", content: trimmedInput }); console.log("\nAgent is thinking...\n"); try { const stream = streamText({ model: openai("gpt-5"), messages, tools, toolChoice: "auto", stopWhen: stepCountIs(10), onStepFinish: (step) => { for (const toolCall of step.toolCalls) { console.log(`[Using tool: ${toolCall.toolName}]`); } if (step.toolCalls.length > 0) { console.log(""); // Add space after tool calls } }, }); for await (const chunk of stream.textStream) { process.stdout.write(chunk); } console.log("\n\n---\n"); // Get final result for message history const response = await stream.response; if (response?.messages?.length) { messages.push(...response.messages); } } catch (error) { console.error("\nAn error occurred while talking to the agent:"); console.error(error); console.log( "\nYou can try again or restart the app if it keeps happening.\n", ); } finally { rl.prompt(); } }); rl.on("close", async () => { await mcpClient.close(); console.log("\n👋 Session ended."); process.exit(0); }); } main().catch((err) => { console.error("Fatal error:", err); process.exit(1); }); ``` ## Complete Code ```typescript import "dotenv/config"; import { openai } from "@ai-sdk/openai"; import { Composio } from "@composio/core"; import * as readline from "readline"; import { streamText, type ModelMessage, stepCountIs } from "ai"; import { createMCPClient } from "@ai-sdk/mcp"; const composioAPIKey = process.env.COMPOSIO_API_KEY; const composioUserID = process.env.COMPOSIO_USER_ID; if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set"); if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set"); if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set"); const composio = new Composio({ apiKey: composioAPIKey, }); async function main() { // Create a tool router session for the user const session = await composio.create(composioUserID!, { toolkits: ["zenrows"], }); const mcpUrl = session.mcp.url; const mcpClient = await createMCPClient({ transport: { type: "http", url: mcpUrl, headers: session.mcp.headers, // Authentication headers for the Composio MCP server }, }); const tools = await mcpClient.tools(); let messages: ModelMessage[] = []; console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n"); console.log( "Ask any questions related to zenrows, like summarize my last 5 emails, send an email, etc... :)))\n", ); const rl = readline.createInterface({ input: process.stdin, output: process.stdout, prompt: "> ", }); rl.prompt(); rl.on("line", async (userInput: string) => { const trimmedInput = userInput.trim(); if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) { console.log("\nGoodbye!"); rl.close(); process.exit(0); } if (!trimmedInput) { rl.prompt(); return; } messages.push({ role: "user", content: trimmedInput }); console.log("\nAgent is thinking...\n"); try { const stream = streamText({ model: openai("gpt-5"), messages, tools, toolChoice: "auto", stopWhen: stepCountIs(10), onStepFinish: (step) => { for (const toolCall of step.toolCalls) { console.log(`[Using tool: ${toolCall.toolName}]`); } if (step.toolCalls.length > 0) { console.log(""); // Add space after tool calls } }, }); for await (const chunk of stream.textStream) { process.stdout.write(chunk); } console.log("\n\n---\n"); // Get final result for message history const response = await stream.response; if (response?.messages?.length) { messages.push(...response.messages); } } catch (error) { console.error("\nAn error occurred while talking to the agent:"); console.error(error); console.log( "\nYou can try again or restart the app if it keeps happening.\n", ); } finally { rl.prompt(); } }); rl.on("close", async () => { await mcpClient.close(); console.log("\n👋 Session ended."); process.exit(0); }); } main().catch((err) => { console.error("Fatal error:", err); process.exit(1); }); ``` ## Conclusion You've successfully built a Zenrows agent using the Vercel AI SDK with streaming capabilities! This implementation provides a powerful foundation for building AI applications with natural language interfaces and real-time feedback. Key features of this implementation: - Real-time streaming responses for a better user experience with typewriter effect - Live tool execution feedback showing which tools are being used as the agent works - Dynamic tool loading through Composio's Tool Router with secure authentication - Multi-step tool execution with configurable step limits (up to 10 steps) - Comprehensive error handling for robust agent execution - Conversation history maintenance for context-aware responses You can extend this further by adding custom error handling, implementing specific business logic, or integrating additional Composio toolkits to create multi-app workflows. ## How to build Zenrows MCP Agent with another framework - [OpenAI Agents SDK](https://composio.dev/toolkits/zenrows/framework/open-ai-agents-sdk) - [Claude Agent SDK](https://composio.dev/toolkits/zenrows/framework/claude-agents-sdk) - [Claude Code](https://composio.dev/toolkits/zenrows/framework/claude-code) - [Claude Cowork](https://composio.dev/toolkits/zenrows/framework/claude-cowork) - [Codex](https://composio.dev/toolkits/zenrows/framework/codex) - [OpenClaw](https://composio.dev/toolkits/zenrows/framework/openclaw) - [Hermes](https://composio.dev/toolkits/zenrows/framework/hermes-agent) - [CLI](https://composio.dev/toolkits/zenrows/framework/cli) - [Google ADK](https://composio.dev/toolkits/zenrows/framework/google-adk) - [LangChain](https://composio.dev/toolkits/zenrows/framework/langchain) - [Mastra AI](https://composio.dev/toolkits/zenrows/framework/mastra-ai) - [LlamaIndex](https://composio.dev/toolkits/zenrows/framework/llama-index) - [CrewAI](https://composio.dev/toolkits/zenrows/framework/crew-ai) ## Related Toolkits - [Supabase](https://composio.dev/toolkits/supabase) - Supabase is an open-source backend platform offering scalable Postgres databases, authentication, storage, and real-time APIs. It lets developers build modern apps without managing infrastructure. - [Codeinterpreter](https://composio.dev/toolkits/codeinterpreter) - Codeinterpreter is a Python-based coding environment with built-in data analysis and visualization. It lets you instantly run scripts, plot results, and prototype solutions inside supported platforms. - [GitHub](https://composio.dev/toolkits/github) - GitHub is a code hosting platform for version control and collaborative software development. It streamlines project management, code review, and team workflows in one place. - [Ably](https://composio.dev/toolkits/ably) - Ably is a real-time messaging platform for live chat and data sync in modern apps. It offers global scale and rock-solid reliability for seamless, instant experiences. - [Abuselpdb](https://composio.dev/toolkits/abuselpdb) - Abuselpdb is a central database for reporting and checking IPs linked to malicious online activity. Use it to quickly identify and report suspicious or abusive IP addresses. - [Alchemy](https://composio.dev/toolkits/alchemy) - Alchemy is a blockchain development platform offering APIs and tools for Ethereum apps. It simplifies building and scaling Web3 projects with robust infrastructure. - [Algolia](https://composio.dev/toolkits/algolia) - Algolia is a hosted search API that powers lightning-fast, relevant search experiences for web and mobile apps. It helps developers deliver instant, typo-tolerant, and scalable search without complex infrastructure. - [Anchor browser](https://composio.dev/toolkits/anchor_browser) - Anchor browser is a developer platform for AI-powered web automation. It transforms complex browser actions into easy API endpoints for streamlined web interaction. - [Apiflash](https://composio.dev/toolkits/apiflash) - Apiflash is a website screenshot API for programmatically capturing web pages. It delivers high-quality screenshots on demand for automation, monitoring, or reporting. - [Apiverve](https://composio.dev/toolkits/apiverve) - Apiverve delivers a suite of powerful APIs that simplify integration for developers. It's designed for reliability and scalability so you can build faster, smarter applications without the integration headache. - [Appcircle](https://composio.dev/toolkits/appcircle) - Appcircle is an enterprise-grade mobile CI/CD platform for building, testing, and publishing mobile apps. It streamlines mobile DevOps so teams ship faster and with more confidence. - [Appdrag](https://composio.dev/toolkits/appdrag) - Appdrag is a cloud platform for building websites, APIs, and databases with drag-and-drop tools and code editing. It accelerates development and iteration by combining hosting, database management, and low-code features in one place. - [Appveyor](https://composio.dev/toolkits/appveyor) - AppVeyor is a cloud-based continuous integration service for building, testing, and deploying applications. It helps developers automate and streamline their software delivery pipelines. - [Backendless](https://composio.dev/toolkits/backendless) - Backendless is a backend-as-a-service platform for mobile and web apps, offering database, file storage, user authentication, and APIs. It helps developers ship scalable applications faster without managing server infrastructure. - [Baserow](https://composio.dev/toolkits/baserow) - Baserow is an open-source no-code database platform for building collaborative data apps. It makes it easy for teams to organize data and automate workflows without writing code. - [Bench](https://composio.dev/toolkits/bench) - Bench is a benchmarking tool for automated performance measurement and analysis. It helps you quickly evaluate, compare, and track your systems or workflows. - [Better stack](https://composio.dev/toolkits/better_stack) - Better Stack is a monitoring, logging, and incident management solution for apps and services. It helps teams ensure application reliability and performance with real-time insights. - [Bitbucket](https://composio.dev/toolkits/bitbucket) - Bitbucket is a Git-based code hosting and collaboration platform for teams. It enables secure repository management and streamlined code reviews. - [Blazemeter](https://composio.dev/toolkits/blazemeter) - Blazemeter is a continuous testing platform for web and mobile app performance. It empowers teams to automate and analyze large-scale tests with ease. - [Blocknative](https://composio.dev/toolkits/blocknative) - Blocknative delivers real-time mempool monitoring and transaction management for public blockchains. Instantly track pending transactions and optimize blockchain interactions with live data. ## Frequently Asked Questions ### What are the differences in Tool Router MCP and Zenrows MCP? With a standalone Zenrows MCP server, the agents and LLMs can only access a fixed set of Zenrows tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Zenrows and many other apps based on the task at hand, all through a single MCP endpoint. ### Can I use Tool Router MCP with Vercel AI SDK v6? Yes, you can. Vercel AI SDK v6 fully supports MCP integration. You get structured tool calling, message history handling, and model orchestration while Tool Router takes care of discovering and serving the right Zenrows tools. ### Can I manage the permissions and scopes for Zenrows while using Tool Router? Yes, absolutely. You can configure which Zenrows scopes and actions are allowed when connecting your account to Composio. You can also bring your own OAuth credentials or API configuration so you keep full control over what the agent can do. ### How safe is my data with Composio Tool Router? All sensitive data such as tokens, keys, and configuration is fully encrypted at rest and in transit. Composio is SOC 2 Type 2 compliant and follows strict security practices so your Zenrows data and credentials are handled as safely as possible. --- [See all toolkits](https://composio.dev/toolkits) · [Composio docs](https://docs.composio.dev/llms.txt)