Content Prompt Maker helps you turn a collection of content decisions into one structured instruction for an LLM. It does not write the final article, video script, or social post, and it does not send your brief to an AI service. Instead, the browser tool combines the settings and context you provide into a prompt that you can review, copy, and use with the LLM of your choice.
This separation is useful when a short request would leave too much unstated. A topic alone does not tell an LLM who the audience is, where the content will be published, what evidence is available, which format is required, or which visual and SEO deliverables belong with the main text. Content Prompt Maker makes those choices visible before generation begins.
Start with the content type and platform
First choose the kind of content you are planning. The available profiles include Blog, YouTube Long-form, Shorts / Reels / TikTok, Instagram, Facebook, Threads, LinkedIn, and Pinterest. Each profile shows settings that are relevant to that workflow instead of presenting one oversized form for every platform.
For a blog, the publishing platform can shape the requested result. Universal / Other provides a general structure, while WordPress, Ghost, Medium, Blogger, Substack, Naver Blog, and Brunch can be used when you want the prompt to acknowledge a particular publishing environment. The platform choice is guidance for the LLM, not a direct connection to a CMS. Wavenox does not sign in to or publish on any of those services.
Add the context the LLM actually needs
A useful production prompt distinguishes the subject from the evidence and point of view behind it. Enter the topic, define the target audience, and add reference material when you have it. You can also describe real user experience, state your own opinion, list details that must be included, identify material that should be excluded, and add any final instructions that do not fit another field.
These fields also establish an important accuracy boundary. If you do not provide personal experience, a source, a quotation, or a statistic, the prompt should not invite the LLM to invent one. Reference material can show what the answer may rely on, while include and exclude requirements make the limits explicit. This cannot prevent every model error, but it gives you a clearer basis for reviewing the result.
Choose the output requirements
Blog workflows can request Plain Text, Markdown, or Clean HTML. You can decide how the title should be handled and whether the result should include SEO metadata, categories, tags, or an FAQ. Visual options can add a featured-image prompt and inline-image prompts, along with ALT text, captions, aspect ratios, and placement suggestions. These are instructions for the LLM to produce a complete content package; they are not images generated by Wavenox.
Choose only the deliverables you expect to use. A compact draft may need a title and body, while a publishing handoff may benefit from metadata, image directions, and a separate FAQ. Removing unnecessary requirements keeps the generated prompt easier to inspect and gives the LLM fewer competing tasks.
Build a YouTube-specific workflow
The YouTube Long-form profile adds planning choices that are different from a blog. You can set a target duration, script type, hook strength, speaking rhythm, production style, on-camera or faceless appearance, available visual resources, editing rhythm, and call to action. Optional deliverables include titles, a description, tags, a thumbnail package, and visual or supporting-footage guidance.
For narration workflows, TTS-ready clean script means the completed narration should not be mixed with timestamps, Markdown headings, chapter markers, production notes, or visual notes. Chapters remain an independent option. With Chapters OFF, the prompt asks for no timestamp or chapter markers anywhere in the result. With Chapters ON and TTS enabled, the clean narration stays in its own section and the Chapters list is requested separately. This makes it easier to copy narration into a voice workflow without manually removing production markup.
Watch the prompt update as settings change
As you change settings on the left, the generated prompt on the right updates immediately. You can therefore compare the effect of a different platform, format, audience, or deliverable before copying anything. The goal is not to claim a perfect prompt; it is to reduce the repetitive work of rebuilding a long production brief from the beginning.
Copy the prompt to the LLM you already use
Review the generated instruction, copy it, and paste it into a compatible LLM such as ChatGPT, Claude, Gemini, or Grok. These names are examples of services that can accept text prompts, not official integrations or endorsements. Content Prompt Maker does not connect to their APIs, submit the prompt for you, or receive the generated result. You remain in control of where the prompt is used.
Save and reload repeatable settings
Content Prompt Maker does not automatically save its settings to localStorage. Use the Save control when you want to create a JSON settings file, and use Load when you want to restore that file later. In supported browsers, the File System Access API can let you choose a save location. Other environments use a normal file-download fallback. This explicit workflow makes it possible to reproduce a recurring content brief without silently storing every draft in the browser.
Review accuracy before publishing
A structured prompt cannot guarantee a correct, original, compliant, or effective final result. Read the LLM output before using it. Verify facts, sources, quotations, statistics, legal or policy claims, and any time-sensitive details when they matter. Even when the prompt tells the model not to invent user experience, the final answer can still contain unsupported details that require human review.
A simple example workflow
Suppose the topic is “How small businesses can use AI for content planning.” Choose Blog as the content type, WordPress as the platform, and small business owners who are new to AI as the audience. Select Clean HTML, then request SEO metadata, an FAQ, and a featured-image prompt. Add any trusted references and note that the answer must not invent case studies or performance statistics.
The prompt panel now combines those decisions into one production instruction. Review it, save the settings as JSON if this will be a recurring workflow, then copy the prompt into your preferred LLM. The LLM produces the draft; you verify the claims, edit the wording, check the HTML and metadata, and publish through your own WordPress process. Wavenox remains the planning layer rather than the generator or publisher.