{"id":46,"date":"2026-04-19T21:14:58","date_gmt":"2026-04-19T21:14:58","guid":{"rendered":"https:\/\/thethriftydev.com\/blog\/how-to-build-your-first-ai-agent-without-code-2026-step-by-step-guide\/"},"modified":"2026-04-24T22:29:55","modified_gmt":"2026-04-24T22:29:55","slug":"how-to-build-your-first-ai-agent-without-code-2026-step-by-step-guide","status":"publish","type":"post","link":"https:\/\/thethriftydev.com\/blog\/how-to-build-your-first-ai-agent-without-code-2026-step-by-step-guide\/","title":{"rendered":"How to Build Your First AI Agent Without Code (2026 Step-by-Step Guide)"},"content":{"rendered":"<h1>How to Build Your First AI Agent Without Code (2026 Step-by-Step Guide)<\/h1>\n<p><strong>You&#8217;ve heard the buzz about AI agents. Everyone from your LinkedIn feed to your techy cousin is talking about them. But every tutorial you find assumes you know Python, LangChain, and have a GitHub account older than your car. What if you just want to build one \u2014 no code required?<\/strong><\/p>\n<p>Good news: you can. In 2026, the tools have finally caught up with the hype. This guide walks you through building a real, working AI agent using nothing but drag-and-drop interfaces and free tools. No terminal, no pip install, no Stack Overflow rabbit holes.<\/p>\n<p>If you&#8217;re already familiar with visual workflow builders (we covered why they&#8217;re the future in <a href=\"https:\/\/thethriftydev.com\/blog\/the-click-not-code-manifesto-why-visual-workflow-builders-are-the-future-3\">The &#8216;Click Not Code&#8217; Manifesto<\/a>), you&#8217;re halfway there. AI agents are the natural next step beyond simple automations.<\/p>\n<\/p>\n<h2>What Is an AI Agent (And Why It&#8217;s Different from a Chatbot or Workflow)<\/h2>\n<p>Let&#8217;s clear this up because the internet keeps mixing these terms up.<\/p>\n<p>A <strong>chatbot<\/strong> answers questions. You ask, it responds. End of story. It&#8217;s reactive \u2014 it waits for you.<\/p>\n<p>A <strong>workflow<\/strong> (the kind you build in Zapier or n8n) follows a fixed path: &#8220;When X happens, do Y, then Z.&#8221; Predictable. Reliable. Dumb. It doesn&#8217;t think \u2014 it follows rules you set.<\/p>\n<p>An <strong>AI agent<\/strong> is different. It has a goal, access to tools, and the ability to <em>decide<\/em> what to do next. Instead of &#8220;if this, then that,&#8221; it&#8217;s &#8220;here&#8217;s my goal, here are my tools \u2014 I&#8217;ll figure out the steps.&#8221;<\/p>\n<p>Think of it this way: a workflow is a recipe. An AI agent is a chef who can improvise when an ingredient is missing. The agent can search the web, read documents, call APIs, and chain multiple actions together based on what it finds \u2014 all without you scripting every step.<\/p>\n<\/p>\n<h2>Why 2026 Is the Year of AI Agents \u2014 What Changed<\/h2>\n<p>AI agents aren&#8217;t new conceptually \u2014 researchers have talked about them for decades. But three things changed in the last year that made them actually usable for regular people:<\/p>\n<ul>\n<li><strong>OpenAI launched the Agents SDK and Responses API<\/strong> \u2014 purpose-built tools for creating agents with built-in web search, file search, and computer use. No more hacking together chat APIs and pretending they&#8217;re agents. (<a href=\"https:\/\/openai.com\/index\/new-tools-for-building-agents\/\" rel=\"noopener noreferrer\">source<\/a>)<\/li>\n<li><strong>No-code platforms added real agent builders<\/strong> \u2014 n8n, Make.com, and others now have visual agent nodes that integrate directly with LLMs. You drag, drop, configure, and deploy.<\/li>\n<li><strong>Models got smart enough to be useful<\/strong> \u2014 earlier LLMs were too unreliable for autonomous tasks. Current models can plan, reason, and recover from errors well enough to trust with real work.<\/li>\n<\/ul>\n<p>IBM called it at their Think 2026 conference: agentic AI is the biggest shift since cloud computing. Major tech companies are already deploying AI agents for cybersecurity, research, and enterprise operations. But you don&#8217;t need a billion-dollar budget to play \u2014 you need a free afternoon.<\/p>\n<\/p>\n<h2>What You Need Before You Start (No Coding Required)<\/h2>\n<p>Before we build, let&#8217;s make sure you have the basics:<\/p>\n<ul>\n<li><strong>An OpenAI API key<\/strong> (or Anthropic, or Google \u2014 pick your poison). Costs a few dollars to experiment. You can get one at <a href=\"https:\/\/platform.openai.com\" rel=\"noopener noreferrer\">platform.openai.com<\/a>.<\/li>\n<li><strong>A free n8n account<\/strong> \u2014 either cloud (n8n.cloud) or self-hosted. The cloud free tier is generous enough for learning.<\/li>\n<li><strong>A clear idea of what your agent should do<\/strong> \u2014 &#8220;customer support for my store&#8221; or &#8220;research competitor pricing&#8221; are great starting points. &#8220;Build me something cool&#8221; is not.<\/li>\n<li><strong>30-60 minutes of focused time<\/strong> \u2014 no half-watching YouTube tutorials while building. Give it your full attention.<\/li>\n<\/ul>\n<p>Already built automations before? Great \u2014 check out our guide on <a href=\"https:\/\/thethriftydev.com\/blog\/building-your-first-ai-workflow-a-complete-beginners-guide-2\">Building Your First AI Workflow<\/a> for context, then come back here to level up to agents.<\/p>\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/thethriftydev.com\/blog\/wp-content\/uploads\/2026\/04\/How_to_Build_Your_Fi.jpg\" alt=\"How to Build Your First AI Agent Without Code (202\" loading=\"lazy\" \/><\/figure>\n<\/p>\n<h2>The Best No-Code AI Agent Builders in 2026 (Compared)<\/h2>\n<p>Here&#8217;s the landscape as of April 2026. I&#8217;ve tested all of these so you don&#8217;t have to waste your weekend:<\/p>\n<table>\n<thead>\n<tr>\n<th>Tool<\/th>\n<th>Free Tier<\/th>\n<th>Best For<\/th>\n<th>Learning Curve<\/th>\n<th>Agent Features<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><strong>n8n<\/strong><\/td>\n<td>Yes (generous)<\/td>\n<td>Complex agents with integrations<\/td>\n<td>Medium<\/td>\n<td>LangChain-based agent nodes, tool calling, memory<\/td>\n<\/tr>\n<tr>\n<td><strong>OpenAI Agent Builder<\/strong><\/td>\n<td>Yes (pay per use)<\/td>\n<td>Simple agents using GPT models<\/td>\n<td>Low<\/td>\n<td>Built-in web search, file search, code interpreter<\/td>\n<\/tr>\n<tr>\n<td><strong>Make.com<\/strong><\/td>\n<td>Yes (limited)<\/td>\n<td>Business process agents<\/td>\n<td>Low-Medium<\/td>\n<td>AI modules with basic agent capabilities<\/td>\n<\/tr>\n<tr>\n<td><strong>Zapier<\/strong><\/td>\n<td>Yes (limited)<\/td>\n<td>Simple trigger-based AI actions<\/td>\n<td>Very Low<\/td>\n<td>AI steps, but not true autonomous agents<\/td>\n<\/tr>\n<tr>\n<td><strong>CrewAI (no-code)<\/strong><\/td>\n<td>Yes<\/td>\n<td>Multi-agent systems<\/td>\n<td>Medium-High<\/td>\n<td>Multiple agents collaborating on tasks<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>For this guide, we&#8217;ll use <strong>n8n<\/strong> (best free option with real agent power) and <strong>OpenAI&#8217;s Agent Builder<\/strong> (easiest to get started). If you want a deeper comparison of automation platforms, we&#8217;ve got you covered with <a href=\"https:\/\/thethriftydev.com\/blog\/n8n-vs-make-com-vs-zapier-which-automation-tool-is-right-for-you-in-2026\">n8n vs Make.com vs Zapier in 2026<\/a>.<\/p>\n<\/p>\n<h2>Step-by-Step: Build a Customer Support AI Agent in n8n (Free)<\/h2>\n<p>This is the part you came for. Let&#8217;s build a customer support agent that can:<\/p>\n<ul>\n<li>Receive incoming support requests (email, form, or chat)<\/li>\n<li>Understand the issue and categorize it<\/li>\n<li>Search your knowledge base for answers<\/li>\n<li>Respond to the customer or escalate to a human<\/li>\n<\/ul>\n<p><strong>Step 1: Create a new workflow in n8n.<\/strong> Log in, click &#8220;Add Workflow.&#8221; You&#8217;ll see the canvas \u2014 your playground.<\/p>\n<p><strong>Step 2: Add a trigger.<\/strong> Click the + button and choose your input method. For this example, use &#8220;Webhook&#8221; (simulates receiving a support ticket). Configure it to accept POST requests.<\/p>\n<p><strong>Step 3: Add the AI Agent node.<\/strong> This is the magic. Search for &#8220;AI Agent&#8221; in the nodes panel. Drag it onto the canvas. n8n&#8217;s agent node is built on LangChain and gives you a visual interface for configuring your agent&#8217;s brain.<\/p>\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/thethriftydev.com\/blog\/wp-content\/uploads\/2026\/04\/fix_How_to_Build_Your_Fi-1.jpg\" alt=\"How to Build Your First AI Agent Without Code (202\" loading=\"lazy\" \/><\/figure>\n<\/p>\n<p><strong>Step 4: Configure your agent.<\/strong> You&#8217;ll need to set three things:<\/p>\n<ul>\n<li><strong>Model:<\/strong> Connect your OpenAI API key and select GPT-4o (or whatever model you prefer).<\/li>\n<li><strong>System prompt:<\/strong> This is your agent&#8217;s personality and instructions. Example: &#8220;You are a customer support agent for [Your Business]. Be helpful, concise, and friendly. If you can&#8217;t answer a question confidently, say so and escalate.&#8221;<\/li>\n<li><strong>Tools:<\/strong> Add a &#8220;Vector Store&#8221; tool (connects to your knowledge base documents) and an &#8220;HTTP Request&#8221; tool (lets the agent look things up).<\/li>\n<\/ul>\n<p><strong>Step 5: Add memory.<\/strong> Attach a &#8220;Window Buffer Memory&#8221; node so the agent remembers the conversation context. Without this, every message starts fresh \u2014 useless for support.<\/p>\n<p><strong>Step 6: Add output actions.<\/strong> After the agent responds, add nodes to send the reply (email, Slack, etc.) and optionally log the conversation in a Google Sheet or database.<\/p>\n<p><strong>Step 7: Test it.<\/strong> Click &#8220;Test Workflow,&#8221; send a sample support request through your webhook, and watch the agent think, search, and respond. It&#8217;s genuinely satisfying the first time it works.<\/p>\n<\/p>\n<h2>Step-by-Step: Build a Research Agent with OpenAI&#8217;s Agent Builder<\/h2>\n<p>Want something even simpler? OpenAI&#8217;s Agent Builder lets you create agents directly in their platform with minimal setup. Here&#8217;s how to build a research agent that gathers and summarizes information on any topic:<\/p>\n<p><strong>Step 1:<\/strong> Go to <a href=\"https:\/\/platform.openai.com\/docs\/guides\/agents\" rel=\"noopener noreferrer\">OpenAI&#8217;s Agents documentation<\/a> and access the Agent Builder interface.<\/p>\n<p><strong>Step 2:<\/strong> Define your agent&#8217;s purpose. For a research agent: &#8220;You are a research assistant. When given a topic, search the web for the latest information, synthesize findings, and provide a structured summary with sources.&#8221;<\/p>\n<p><strong>Step 3:<\/strong> Enable tools \u2014 specifically <strong>web search<\/strong> and <strong>file search<\/strong>. These give your agent the ability to actually look things up rather than relying on training data.<\/p>\n<p><strong>Step 4:<\/strong> Set guardrails. Tell the agent what it should NOT do (no making up sources, no going off-topic, always cite where info came from).<\/p>\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/thethriftydev.com\/blog\/wp-content\/uploads\/2026\/04\/How_to_Build_Your_Fi-2.jpg\" alt=\"How to Build Your First AI Agent Without Code (202\" loading=\"lazy\" \/><\/figure>\n<\/p>\n<p><strong>Step 5:<\/strong> Test with real queries. Try &#8220;What are the latest developments in AI agent frameworks?&#8221; and see what comes back. Iterate on your instructions until the output matches what you need.<\/p>\n<p>The beauty of this approach is speed \u2014 you can have a working research agent in under 15 minutes. The tradeoff is flexibility; n8n gives you way more control over integrations and custom logic. <em>Note: OpenAI&#8217;s interface evolves quickly \u2014 the exact steps may differ slightly from what you see, but the core concepts remain the same.<\/em><\/p>\n<\/p>\n<h2>Common Mistakes Beginners Make with AI Agents<\/h2>\n<p>I&#8217;ve seen these mistakes burn countless beginners (and honestly, I made most of them myself):<\/p>\n<ul>\n<li><strong>Vague instructions.<\/strong> &#8220;Be helpful&#8221; is not a system prompt. Be specific: &#8220;You are a support agent for a SaaS tool that helps small businesses with invoicing. Answer billing questions, troubleshoot login issues, and escalate anything about API integration.&#8221;<\/li>\n<li><strong>Too many tools.<\/strong> Giving an agent 15 tools sounds powerful. In practice, it confuses the model. Start with 2-3 tools and add more only when needed.<\/li>\n<li><strong>No guardrails.<\/strong> Without clear boundaries, agents will happily make up answers, share sensitive info, or go down rabbit holes. Always include &#8220;if unsure, say so&#8221; in your instructions.<\/li>\n<li><strong>Skipping the memory.<\/strong> An agent without memory is like a goldfish. Every interaction starts from zero. Add conversation memory from day one.<\/li>\n<li><strong>Not testing edge cases.<\/strong> Your agent handles normal queries fine. But what about angry customers? Gibberish input? Questions outside its scope? Test the weird stuff.<\/li>\n<\/ul>\n<h2>When to Use an AI Agent vs a Simple Automation Workflow<\/h2>\n<p>Not everything needs an agent. Sometimes a plain workflow is better \u2014 cheaper, faster, more predictable. Here&#8217;s how to decide:<\/p>\n<p><strong>Use a simple workflow when:<\/strong><\/p>\n<ul>\n<li>The path is always the same (if X, then Y)<\/li>\n<li>You need 100% reliability and predictability<\/li>\n<li>The task doesn&#8217;t require understanding natural language<\/li>\n<li>Speed matters more than flexibility<\/li>\n<\/ul>\n<p><strong>Use an AI agent when:<\/strong><\/p>\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/thethriftydev.com\/blog\/wp-content\/uploads\/2026\/04\/How_to_Build_Your_Fi-3.jpg\" alt=\"How to Build Your First AI Agent Without Code (202\" loading=\"lazy\" \/><\/figure>\n<\/p>\n<ul>\n<li>The task requires understanding unstructured input (emails, chat messages, documents)<\/li>\n<li>There are multiple possible paths based on context<\/li>\n<li>The agent needs to search for information before acting<\/li>\n<li>Human-like judgment is needed (triaging, categorizing, deciding)<\/li>\n<\/ul>\n<h2>How to Test and Improve Your AI Agent<\/h2>\n<p>Building the agent is 20% of the work. The other 80% is making it not suck. Here&#8217;s your testing playbook:<\/p>\n<p><strong>1. Start with golden examples.<\/strong> Write 10-15 test inputs that represent your ideal use cases. Run them all through the agent and check the outputs. Fix instructions based on failures.<\/p>\n<p><strong>2. Add adversarial tests.<\/strong> Now throw the worst inputs you can think of at it: gibberish, edge cases, hostile messages, requests outside scope. Your agent should handle all of these gracefully.<\/p>\n<p><strong>3. Log everything.<\/strong> Use n8n&#8217;s execution logs (or OpenAI&#8217;s built-in tracing) to see exactly what your agent did at each step. This is how you find where things go wrong.<\/p>\n<p><strong>4. Iterate on the system prompt.<\/strong> Small wording changes have outsized effects. &#8220;Be concise&#8221; vs &#8220;Answer in 2-3 sentences maximum&#8221; produces very different results. Be specific, test, repeat.<\/p>\n<p><strong>5. Monitor in production.<\/strong> Once your agent is live, check the first 50-100 real interactions manually. You&#8217;ll find patterns the agent gets wrong and can adjust before they become real problems.<\/p>\n<\/p>\n<h2>What&#8217;s Next: Multi-Agent Systems and the Future<\/h2>\n<p>Once you&#8217;ve built one agent, the next frontier is <strong>multi-agent systems<\/strong> \u2014 multiple agents working together. Think of it as building a team instead of hiring one person.<\/p>\n<p>OpenAI&#8217;s Agents SDK now supports multi-agent orchestration out of the box. CrewAI&#8217;s no-code interface lets you define agents with different roles (researcher, writer, reviewer) that collaborate on complex tasks. Forbes recently reported that reinforcement learning gyms are using real workplace data to train agents to navigate actual business environments \u2014 we&#8217;re moving fast.<\/p>\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/thethriftydev.com\/blog\/wp-content\/uploads\/2026\/04\/How_to_Build_Your_Fi-4.jpg\" alt=\"How to Build Your First AI Agent Without Code (202\" loading=\"lazy\" \/><\/figure>\n<\/p>\n<p>But don&#8217;t get ahead of yourself. Master a single agent first. Understand how prompts, tools, and memory interact. Then graduate to multi-agent setups. Crawl, walk, run.<\/p>\n<p>The tools are only getting better. Prices are dropping. And the barrier to entry is lower than ever. If you&#8217;ve been waiting for the &#8220;right time&#8221; to start building AI agents \u2014 this is it.<\/p>\n<\/p>\n<h2>FAQ<\/h2>\n<p><strong>Q: Do I really need zero coding experience to build an AI agent?<\/strong><br \/>\nA: Yes, genuinely. Tools like n8n and OpenAI&#8217;s Agent Builder are visual and configuration-driven. You need to be comfortable with APIs (copy-pasting a key) and logical thinking, but not a single line of code. If you can set up a Zapier workflow, you can build an AI agent.<\/p>\n<p><strong>Q: How much does it cost to run an AI agent?<\/strong><br \/>\nA: For learning and light use: under $5\/month. You&#8217;ll pay for the LLM API calls (OpenAI charges per token), but a single agent handling a few dozen interactions a day costs pennies. n8n&#8217;s free tier handles plenty of volume. As you scale up, costs increase \u2014 but by then, the agent should be saving you money elsewhere.<\/p>\n<p><strong>Q: What&#8217;s the difference between an AI agent and a regular chatbot like ChatGPT?<\/strong><br \/>\nA: ChatGPT is a conversational AI \u2014 you chat, it responds. An AI agent has <em>goals and tools<\/em>. It can take actions autonomously: search the web, read files, send emails, update databases, call APIs. A chatbot talks. An agent <em>does<\/em>.<\/p>\n<p><strong>Q: Can I use these agents for my business immediately?<\/strong><br \/>\nA: You can prototype quickly, but production use requires testing. Start with internal use cases (research, data processing) before customer-facing deployments. Add guardrails, monitor outputs, and always have a human in the loop for critical decisions.<\/p>\n<hr>\n<p><em>Sources: <a href=\"https:\/\/openai.com\/index\/new-tools-for-building-agents\/\" rel=\"noopener noreferrer\">OpenAI Blog \u2014 New Tools for Building Agents<\/a> \u00b7 <a href=\"https:\/\/blog.n8n.io\/ai-agents\/\" rel=\"noopener noreferrer\">n8n Blog \u2014 AI Agents Explained<\/a> \u00b7 <a href=\"https:\/\/platform.openai.com\/docs\/guides\/agents\" rel=\"noopener noreferrer\">OpenAI Agents SDK Documentation<\/a> \u00b7 <a href=\"https:\/\/www.ibm.com\/think\" rel=\"noopener noreferrer\">IBM Think \u2014 AI Agents Insights<\/a><\/em><\/p>\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/thethriftydev.com\/blog\/wp-content\/uploads\/2026\/04\/How_to_Build_Your_Fi-5.jpg\" alt=\"How to Build Your First AI Agent Without Code (202\" loading=\"lazy\" \/><\/figure>\n<\/p>\n<p>Follow <a href=\"https:\/\/x.com\/TheThriftyDev\" rel=\"noopener noreferrer\">@TheThriftyDev<\/a> for more no-code AI tutorials, automation tips, and straight talk about building things without breaking the bank.<\/p>\n<p>Related reads:<br \/>\n\u2192 <a href=\"https:\/\/thethriftydev.com\/blog\/the-click-not-code-manifesto-why-visual-workflow-builders-are-the-future-3\">The &#8216;Click Not Code&#8217; Manifesto: Why Visual Workflow Builders Are the Future<\/a><br \/>\n\u2192 <a href=\"https:\/\/thethriftydev.com\/blog\/building-your-first-ai-workflow-a-complete-beginners-guide-2\">Building Your First AI Workflow: A Complete Beginner&#8217;s Guide<\/a><br \/>\n\u2192 <a href=\"https:\/\/thethriftydev.com\/blog\/n8n-vs-make-com-vs-zapier-which-automation-tool-is-right-for-you-in-2026\">n8n vs Make.com vs Zapier: Which Automation Tool Is Right for You in 2026<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>How to Build Your First AI Agent Without Code (2026 Step-by-Step Guide) You&#8217;ve heard the buzz about AI agents. Everyone from your LinkedIn feed to your techy cousin is talking about them. But every tutorial you find assumes you know Python, LangChain, and have a GitHub account older than your car. What if you just&hellip; <a class=\"more-link\" href=\"https:\/\/thethriftydev.com\/blog\/how-to-build-your-first-ai-agent-without-code-2026-step-by-step-guide\/\">Continue reading <span class=\"screen-reader-text\">How to Build Your First AI Agent Without Code (2026 Step-by-Step Guide)<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":173,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[],"class_list":["post-46","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-strategy-mindset","entry"],"_links":{"self":[{"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/posts\/46","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/comments?post=46"}],"version-history":[{"count":6,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/posts\/46\/revisions"}],"predecessor-version":[{"id":246,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/posts\/46\/revisions\/246"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/media\/173"}],"wp:attachment":[{"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/media?parent=46"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/categories?post=46"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/thethriftydev.com\/blog\/wp-json\/wp\/v2\/tags?post=46"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}