{"id":55735,"date":"2026-09-04T09:08:31","date_gmt":"2026-09-04T09:08:31","guid":{"rendered":"https:\/\/sitemate.com\/?p=55735"},"modified":"2026-11-12T07:28:14","modified_gmt":"2026-11-12T07:28:14","slug":"how-to-use-chatgpt-to-draft-rams","status":"publish","type":"post","link":"https:\/\/sitemate.com\/za\/mcp\/use-cases\/how-to-use-chatgpt-to-draft-rams\/","title":{"rendered":"How to use ChatGPT to draft RAMS"},"content":{"rendered":"","protected":false},"excerpt":{"rendered":"","protected":false},"author":22,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":"","rank_math_title":"","rank_math_description":"Describe the task, site and crew, and ChatGPT drafts a RAMS: the hazards, a risk score on each, the controls, the method steps in order and a sign-on. You review it, then run and capture it in the field.","rank_math_canonical_url":"","rank_math_focus_keyword":""},"categories":[1],"tags":[],"industry":[],"product":[],"region":[145],"system":[87],"content_type":[170,173],"topic":[157],"class_list":["post-55735","post","type-post","status-publish","format-standard","hentry","category-uncategorized","region-global","system-safety","content_type-ai-landing-page","content_type-llm-use-case-page","topic-rams"],"acf":{"ai_lp_breadcrumb":"<a href=\"\/_hl-staging\/topics\/safety\/rams\/\">RAMS<\/a><span class=\"sep\">\/<\/span><span aria-current=\"page\">How to Use ChatGPT to Draft Rams<\/span>","ai_lp_h1":"How to use ChatGPT to draft RAMS","ai_lp_intro":"A Risk Assessment Method Statement (RAMS) is necessary in all industries, particularly high-risk ones such as construction. This document is essential to verify that safety measures are taken for each step in every activity. However, constructing a RAMS using conventional methods is labor-intensive. RAMS for activities with several steps will require several pages. Creating one would need a significant amount of time and resources. In addition, specific steps contain recurring risks and controls that must be retyped.\nTechnology has provided a tool that can simplify the process of making and completing a RAMS. AI tools, such as ChatGPT, can make a workable RAMS draft in a few minutes. Users only need to describe the specifics of the task in the prompt, and ChatGPT can provide a RAMS, complete with hazards, possible risk scores, suggested safety measures, and a method statement. Safety managers will only need to review the RAMS and tighten it further to make it an official RAMS document.","ai_lp_hero_media":"<div class=\"vimeo-container ailp-video\" data-ailp-video-id=\"1195269765\" data-ailp-video-hash=\"a435d5c75d\" data-ailp-video-title=\"How to use ChatGPT to draft RAMS\" style=\"padding:56.25% 0 0 0;position:relative;\"><\/div>","ai_lp_body":"<div class=\"ailp-block ailp-answer\">\n<p class=\"eyebrow-sm\">The short answer<\/p>\n<p>Provide ChatGPT with the work, site, and crew details, and it will develop a draft RAMS that includes risks, risk scores, suggested safety measures, and a method statement. However, safety managers should review the document before accepting it to verify that there are no gaps, particularly in the required safety measures. Users can also reuse the prompt by modifying the activity characteristics.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>How to get ChatGPT drafting your RAMS<\/h2>\n<p>You don't need a long prompt to get a usable RAMS from ChatGPT. Give it the specifics of the task, tell it the shape a RAMS takes, then tighten what comes back. Once that works, it's a pattern you reuse on the next job, and only the detail of the work changes.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-step-list\">\n<ol class=\"ailp-steps\">\n  <li>\n    <p class=\"eyebrow-sm\">Step 1<\/p>\n    <h3>Give ChatGPT the task and the site<\/h3>\n    <p>The draft is only as specific as what you give ChatGPT. Name the task, the location, the crew and the plant, then say what makes this job different: work at height, a confined space, a live road alongside, a chase cut into a load-bearing wall. That detail is what makes it a RAMS for this job.<\/p>\n    <ul class=\"checklist-tick\">\n      <li>The task, and why it carries risk<\/li>\n      <li>The site and crew, so controls fit the job<\/li>\n      <li>Plant, access and any permits involved<\/li>\n      <li>Anything unusual about this particular job<\/li>\n    <\/ul>\n  <\/li>\n  <li>\n    <p class=\"eyebrow-sm\">Step 2<\/p>\n    <h3>Tell it the shape a RAMS needs<\/h3>\n    <p>Left unprompted, ChatGPT writes prose. Spell out the RAMS sections: hazards, a risk score for each, the controls, the method in sequence and a sign-on line. Match them to the RAMS template your client already accepts, and the draft drops straight in.<\/p>\n    <ul class=\"checklist-tick\">\n      <li>Hazards specific to the task in front of you<\/li>\n      <li>A risk score and residual risk for each<\/li>\n      <li>Controls tied to each hazard<\/li>\n      <li>Method steps in the order the work happens<\/li>\n      <li>A sign-on block for the crew<\/li>\n    <\/ul>\n  <\/li>\n  <li>\n    <p class=\"eyebrow-sm\">Step 3<\/p>\n    <h3>Tighten it to the site<\/h3>\n    <p>Treat the first draft as a starting point. Cut anything generic and add the hazard that's actually on site today. Then check the method against how the crew will really do the work. A few minutes of review is usually what gets it accepted by the principal contractor.<\/p>\n  <\/li>\n<\/ol>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>Prompt guidance for drafting RAMS with ChatGPT<\/h2>\n<p>The prompt is the thing worth building. Write one drafting prompt you can reuse, and change only the task and site each time. The more specific the work you describe, the less generic the RAMS, so every prompt should name the plant and the site as well as the task.<\/p>\n<ul class=\"checklist-dot\">\n  <li>&quot;Draft a RAMS for replacing worn roof sheets on a two-level warehouse. The crew works from a scissor lift with sharp hand tools, and there's an overhead line along the northern boundary.&quot;<\/li>\n  <li>&quot;Draft a RAMS for a welding repair on a leaking water line in a confined space 6 feet high. Score each hazard against the risk matrix below.&quot;<\/li>\n  <li>&quot;Draft a RAMS for a bearing replacement on a 6x4 centrifugal pump used for overflow return. Put the method steps in the order the crew will do them.&quot;<\/li>\n  <li>&quot;Draft a RAMS for erecting scaffolds for a bungalow repair near the beach.&quot;<\/li>\n  <li>&quot;Draft a RAMS for powering a rotary-screw air compressor on and off. Output pressure is 7.5-1.3 bar and working power is 7.5-550 hp. Tie each control to a hazard.&quot;<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>Troubleshooting prompts for drafting RAMS<\/h2>\n<p>A weak draft usually comes from a thin prompt or a template without RAMS structure, rarely from ChatGPT itself. Work through these fixes in order.<\/p>\n<ul class=\"checklist-dot\">\n  <li>Draft too broad: the prompt gave no activity location. Add where the work happens and describe the setting.<\/li>\n  <li>Hazards missing: the prompt didn't name the equipment. List the plant and tools being used.<\/li>\n  <li>Controls too thin: the brief lacked safety input. Bring in competent safety personnel to set the controls.<\/li>\n  <li>Wrong standards cited: no jurisdiction given. State the country in the prompt so the draft follows regional standards.<\/li>\n  <li>Risk ratings look off: the template has no risk matrix. Update the template before generating again.<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>Limitations<\/h2>\n<p>While the use of ChatGPT helps users draft the RAMS, it has limitations on what it can do. These limitations affect the results of the prompt, so reviewing the drafted RAMS is essential.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-subheading-body\">\n<h3>What ChatGPT does well here<\/h3>\n<p>Here are the things that ChatGPT is capable of and can do without any problem:<\/p>\n<ul class=\"checklist-tick\">\n  <li>drafts from your prompts<\/li>\n  <li>tailored to your specific job<\/li>\n  <li>suggests safety controls<\/li>\n  <li>scores hazards against the matrix you provide<\/li>\n  <li>reworks the draft immediately on request<\/li>\n  <li>outputs a Word or Excel document<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-subheading-body\">\n<h3>What it does not do<\/h3>\n<p>These are the limitations of ChatGPT and what users need to be vigilant about:<\/p>\n<ul class=\"checklist-cross\">\n  <li>doesn't know your site environment<\/li>\n  <li>can't draft from conditions on site<\/li>\n  <li>can't guarantee safety measures will work<\/li>\n  <li>can't guarantee a site-ready or audit-ready RAMS<\/li>\n  <li>can't sign off the RAMS<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>What this assumes you already do<\/h2>\n<p>This assumes your team already produces and signs RAMS for its work. ChatGPT drafts the document, so you'll need a habit of reviewing every draft before it's used, and a way of capturing who signed on. Both are part of how the job already runs.<\/p>\n<ul class=\"checklist-tick\">\n  <li>Scope of works and activities are clear<\/li>\n  <li>Project documents and site requirements are to hand<\/li>\n  <li>Hazards and controls are identified<\/li>\n  <li>You know the operatives, tools and plant involved<\/li>\n  <li>You work to the safety procedures that apply<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>Why use ChatGPT to draft RAMS<\/h2>\n<p>A RAMS is not really difficult to make. However, making a RAMS traditionally involves many steps and requires a substantial amount of time and resources. Imagine starting from a master template with 70 pages and having to manually type all identified hazards and safety controls. As a result, companies would resort to just copy-pasting the RAMS, which compromises the needed measures of each specific activity. RAMS like these are usually rejected because of poor quality, which raises serious concerns for auditors.<br>\nUsing modern methods, such as employing AI tools like ChatGPT, can help create a RAMS draft in just a few minutes. Users would only need to provide the specifics of the activity in the prompt, and ChatGPT can generate a RAMS tailored to that activity. Users would now only need to review the RAMS and continue to tighten it to fully fit the needed safety measures for the activity.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>What a good RAMS contains<\/h2>\n<p>These are the criteria a principal contractor checks before accepting a RAMS. A good draft meets all of them.<\/p>\n<ul class=\"checklist-dot\">\n  <li>Hazards considered at every phase of the activity, start to finish<\/li>\n  <li>A full risk assessment for each hazard it identifies<\/li>\n  <li>Safety measures chosen by working down the hierarchy of control<\/li>\n  <li>Every control carried into the method, step by step<\/li>\n  <li>Sign-off by competent, qualified people<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-callout\">\n<p>It's the morning of the job, the principal contractor won't let the crew on site without a signed RAMS, and the one you have is a generic master that doesn't mention the work. Written from scratch now, it holds everyone up. Drafted in minutes and tightened, it doesn't.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-chart\">\n<h2>What people actually run through ChatGPT<\/h2>\n<p>Every named job customers ran through ChatGPT against their own Sitemate data over six months. Close to half the tasks are analysis: counting, reconciling or auditing records that already exist.<\/p>\n<figure class=\"ailp-chart-figure ailp-chart-bar\" data-ailp-chart data-ailp-style=\"bar\">\n<table class=\"ailp-chart-table\">\n<caption class=\"ailp-chart-caption\">What people actually run through ChatGPT<\/caption>\n<thead>\n<tr><th scope=\"col\">The job<\/th><th scope=\"col\">Share of tasks<\/th><\/tr>\n<\/thead>\n<tbody>\n<tr><th scope=\"row\">Users and access<\/th><td data-ailp-value=\"42\">42%<\/td><\/tr>\n<tr><th scope=\"row\">Site diaries<\/th><td data-ailp-value=\"12\">12%<\/td><\/tr>\n<tr><th scope=\"row\">Inspections and checklists<\/th><td data-ailp-value=\"11\">11%<\/td><\/tr>\n<tr><th scope=\"row\">Plant, equipment and fleet<\/th><td data-ailp-value=\"8\">8%<\/td><\/tr>\n<tr><th scope=\"row\">Timesheets and attendance<\/th><td data-ailp-value=\"8\">8%<\/td><\/tr>\n<tr><th scope=\"row\">Invoicing and commercial<\/th><td data-ailp-value=\"7\">7%<\/td><\/tr>\n<tr><th scope=\"row\">Actions and corrective<\/th><td data-ailp-value=\"6\">6%<\/td><\/tr>\n<tr><th scope=\"row\">Compliance and audit<\/th><td data-ailp-value=\"5\">5%<\/td><\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n<\/div>\n\n<div class=\"ailp-block ailp-subheading-body\">\n<h3>Teams reach for the assistant on the documents they repeat<\/h3>\n<p>Teams use assistants most on the documents they repeat: diaries, timesheets, inspections. Each has a fixed structure with details that change from one job to the next. RAMS has the same shape, so drafting one is where ChatGPT earns its place.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>Connecting Sitemate MCP to ChatGPT<\/h2>\n<p>There are three steps, and the first happens outside ChatGPT: an Org Controller turns on MCP access in Dashpivot by Sitemate. You'll need ChatGPT Plus, Team or Enterprise, because connectors aren't available on the free plan.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-step-list\">\n<ol class=\"ailp-steps\">\n  <li>\n    <p class=\"eyebrow-sm\">Step 1<\/p>\n    <h3>Turn on MCP access in Dashpivot<\/h3>\n    <p>An Org Controller goes to Settings > Security in Dashpivot, sets the MCP Access toggle to anything other than None and saves the changes. The app won't authenticate until this is on.<\/p>\n  <\/li>\n  <li>\n    <p class=\"eyebrow-sm\">Step 2<\/p>\n    <h3>Add the Sitemate app in ChatGPT<\/h3>\n    <p>In ChatGPT, go to Settings > Apps and Connections, then Explore Apps. Search for Sitemate, open it and click Connect. Sign in with your Sitemate credentials and approve the permissions.<\/p>\n  <\/li>\n  <li>\n    <p class=\"eyebrow-sm\">Step 3<\/p>\n    <h3>Check the tools are there<\/h3>\n    <p>Open a new chat and confirm the Sitemate tools show under the tools icon. Sitemate MCP is currently read-only: it can read your records but can't create, edit or delete them.<\/p>\n  <\/li>\n<\/ol>\n<\/div>\n\n<div class=\"ailp-block ailp-subheading-body\">\n<h3>If the connector will not appear<\/h3>\n<p>Three causes account for most failed connections, and the first is the most common by far.<\/p>\n<ul class=\"checklist-cross\">\n  <li>Nothing connects: MCP Access is still off in Dashpivot, and only an Org Controller can turn it on<\/li>\n  <li>Sitemate isn't listed: the account needs ChatGPT Plus, Team or Enterprise, as it's not on the free plan<\/li>\n  <li>Tools don't show: start a new chat and check under the tools icon<\/li>\n<\/ul>\n<\/div>\n\n<div class=\"ailp-block ailp-heading-body\">\n<h2>How the record gets captured<\/h2>\n<p>ChatGPT drafts the RAMS. Signing it and keeping it happens in Dashpivot by Sitemate, where the draft becomes a form the crew signs on to. The signed record matters as much as the draft.<\/p>\n<\/div>\n\n<div class=\"ailp-block ailp-step-list\">\n<ol class=\"ailp-steps\">\n  <li>\n    <h3>Storm imports the draft, then captures it on site<\/h3>\n    <p>Save the ChatGPT draft as a PDF, Word or Excel file, and Storm rebuilds it as a Dashpivot template with AI Templates. The RAMS then gets run and signed on a phone or tablet with the crew standing there, and the sign-on is attached. It replaces the emailed copy that gets chased for signatures later.<\/p>\n  <\/li>\n  <li>\n    <h3>Dashpivot holds it as the record<\/h3>\n    <p>Dashpivot holds the RAMS as the record, landed against its job. It's structured and versioned, so it's ready for audit, which a file in someone's inbox isn't.<\/p>\n  <\/li>\n  <li>\n    <h3>Sitemate MCP reads it back<\/h3>\n    <p>Sitemate MCP lets ChatGPT read the captured RAMS, so you can ask which jobs have a signed RAMS and which crews haven't signed on yet. Your existing Dashpivot permissions still apply.<\/p>\n  <\/li>\n<\/ol>\n<\/div>","ai_lp_body_2":"","ai_lp_comparison":"","ai_lp_band_logo":"<img class=\"product-logo\" src=\"https:\/\/sitemate-headless.pages.dev\/shared\/assets\/logos\/dashpivot-logo.svg\" alt=\"Dashpivot\">","ai_lp_band_heading":"Storm captures the RAMS ChatGPT drafts","ai_lp_band_intro":"The crew signs on to the RAMS on site, as the work starts, and Storm captures it there. By the time anyone goes looking for the signed copy, it's already on record.","ai_lp_capabilities":"<div class=\"custom-col-12 bento-col\">\n  <div class=\"custom-grid\">\n    <div class=\"custom-col-6 ailp-cap ailp-cap--split\">\n      <div class=\"ailp-cap-icon\"><img src=\"\/wp-content\/uploads\/2026\/08\/ai-storm-icon.svg\" alt=\"\"><\/div>\n      <div class=\"ailp-cap-text\">\n        <h3>AI Form Fill<\/h3>\n        <p>Capture what you see and say, and Storm drafts the form for you. The details land in the right fields, so there is a record to check the moment the work is done, and you review and submit it yourself.<\/p>\n      <div class=\"button-group\">\n        <a class=\"text-buttons\" href=\"\/storm\/\">Learn more about Storm<img class=\"btn-arrow\" src=\"https:\/\/sitemate-headless.pages.dev\/shared\/assets\/icons\/grey-arrow.svg\" alt=\"\" width=\"16\" height=\"16\"><\/a>\n      <\/div>\n      <\/div>\n    <\/div>\n    <div class=\"custom-col-6\">\n      <img class=\"ailp-media\" src=\"\/wp-content\/uploads\/2026\/08\/storm-form-fill.webp\" alt=\"AI Form Fill\">\n    <\/div>\n  <\/div>\n<\/div>\n\n<div class=\"custom-col-4 bento-col ailp-cap\">\n  <div class=\"ailp-cap-icon\">\n    <div class=\"sm-icon-wrap gradient-light\"><img src=\"\/wp-content\/uploads\/2026\/08\/storm-listens-to-the-work.svg\" alt=\"\"><\/div>\n  <\/div>\n  <div class=\"ailp-cap-text\">\n    <h3>Listens to the work<\/h3>\n    <p>Captures voice cleanly through machinery, wind, and movement. Accents and trade language understood. No headset needed.<\/p>\n  <\/div>\n<\/div>\n\n<div class=\"custom-col-4 bento-col ailp-cap ailp-cap--flip\">\n  <div class=\"ailp-cap-icon\">\n    <div class=\"sm-icon-wrap gradient-light\"><img src=\"\/wp-content\/uploads\/2026\/08\/storm-works-on-site.svg\" alt=\"\"><\/div>\n  <\/div>\n  <div class=\"ailp-cap-text\">\n    <h3>Works on site<\/h3>\n    <p>Built for gloves, cabs, and screen protectors. Captures inputs offline and processes in the background. No waiting on the field.<\/p>\n  <\/div>\n<\/div>\n\n<div class=\"custom-col-4 bento-col ailp-cap\">\n  <div class=\"ailp-cap-icon\">\n    <div class=\"sm-icon-wrap gradient-light\"><img src=\"\/wp-content\/uploads\/2026\/08\/storm-you-stay-in-control.svg\" alt=\"\"><\/div>\n  <\/div>\n  <div class=\"ailp-cap-text\">\n    <h3>You stay in control<\/h3>\n    <p>Storm never commits a record on its own. Every output is a draft for human review. Admins control what Storm can write to.<\/p>\n  <\/div>\n<\/div>","ai_lp_faq_heading":"Frequently Asked Questions","ai_lp_faq":[{"question":"How does ChatGPT identify hazards?","content":"<p>ChatGPT will identify hazards based on what information the user prompts. However, it must be noted that giving a generic prompt will generate a vague RAMS. Users must specify the task, site, equipment involved, and other important aspects of the activity.<\/p>"},{"question":"Will the RAMS generated from ChatGPT be directly usable?","content":"<p>While ChatGPT provides a reasonable draft, it would still be important to conduct a thorough review of the generated RAMS. Reviewing the RAMS can help cut generic content, add real on-site hazards, and ensure a competent person signs off on them.<\/p>"},{"question":"Can ChatGPT use our template?","content":"<p>Yes. Users can upload their existing RAMS template and ask ChatGPT to fill it out based on the prompt the user provides.<\/p>"},{"question":"Where are RAMS in Dashpivot signed and stored?","content":"<p>Users can sign RAMS directly in the field. The RAMS is stored in a cloud storage location and will also be distributed directly to personnel who need to be notified.<\/p>"},{"question":"How is the completed RAMS template distributed?","content":"<p>The Dashpivot platform allows users to modify notifications and send completed forms to relevant personnel who need the RAMS. Once the crew signs a RAMS on site, Storm automatically sends it to relevant stakeholders. Additionally, these completed forms are already versioned and audit-ready.<\/p>"},{"question":"Why do we need a competent person to review the RAMS?","content":"<p>Reviewing the RAMS requires a competent person because these personnel have sufficient knowledge when it comes to providing effective and adequate measures to fully control hazards.<\/p>"}],"ai_lp_version_history":"","ai_lp_closing_heading":"Stop writing every RAMS from a blank page","ai_lp_closing_body":""},"custom_permalink":"mcp\/use-cases\/how-to-use-chatgpt-to-draft-rams","_links":{"self":[{"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/posts\/55735","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/users\/22"}],"replies":[{"embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/comments?post=55735"}],"version-history":[{"count":1,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/posts\/55735\/revisions"}],"predecessor-version":[{"id":55736,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/posts\/55735\/revisions\/55736"}],"wp:attachment":[{"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/media?parent=55735"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/categories?post=55735"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/tags?post=55735"},{"taxonomy":"industry","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/industry?post=55735"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/product?post=55735"},{"taxonomy":"region","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/region?post=55735"},{"taxonomy":"system","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/system?post=55735"},{"taxonomy":"content_type","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/content_type?post=55735"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/sitemate.com\/za\/wp-json\/wp\/v2\/topic?post=55735"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}