SMM builder
Build structured JSON prompts for social media content
Configure post format, platform, engagement, CTA, emojis, brand voice, and extra instructions without manually composing the schema.
Social
SMM
Create an SEO article about choosing a CRM for a small business. Include structure, FAQ, and a clear CTA.
- Common fields + category settings + humanization
{
"category": "smm",
"task": "...",
"common_settings": { ... },
"category_settings": { ... }
}Builder
Builder: SMM
Start with the common settings, then tune the category-specific fields and generate a ready-to-use JSON prompt.
Builder mode
Builder mode
Play mode keeps the whole form on one screen, while Wizard guides the user through clear numbered steps.
Blank mode
Start from scratch
Use it when the request does not fit one category cleanly and you want a more universal JSON prompt setup.
Quick start and presets
Use a built-in scenario to fill the form fast, then export, import, or share your own preset configuration.
Preset families
Core starters
Reliable starting points for the most common use cases in this category.
Instagram engagement
A quick start for a post built to drive comments and saves.
Use case: General use case
Reels soft sale
For a short Reels caption with a benefit-led angle and a soft DM CTA.
Use case: General use case
Telegram expert post
Designed for expert Telegram channels when you need useful, clear copy without hard selling.
Use case: General use case
Task
Describe the plain-language request in your own words. This value will be stored as the task field in the final JSON.
Describe the plain-language request in your own words. This value will be stored as the task field in the final JSON.
Common settings
These parameters affect the result regardless of the selected category.
Who the content or prompt is primarily written for.
Beginners: an audience with limited experience that needs a clear explanation.
How the model should approach the task before producing the final output.
Standard: a direct answer without an extra reasoning framework.
Which language the model should use in the generated result.
English: the output should be written in English.
How concise or detailed the final output should be.
How the final output should sound emotionally.
Friendly: warm, light, and easy to understand.
How the final output should be organized and presented.
Concise: minimal fluff and maximum substance.
In what format the model should return the result.
HTML: the result should be returned in HTML.
Category-specific settings
These fields depend on the selected category and its use case.
Who the model should act like while writing.
Expert copywriter: writes clearly, persuasively, and to the point.
What type of content the model should create.
Instagram post: a standard text for the feed.
Where the text will be published.
Instagram: best for visual, emotional, easy-to-read posts.
What result you want to get from the post.
Engagement: you want comments, reactions, and discussion.
What the text will be about.
Meditation: calmness, focus, and stress reduction.
What form the result should take.
Plain text: just the text without extra visual elements.
What structure the post should follow.
Hook -> body -> CTA: first grab attention, then explain, then ask for action.
What kind of first line should be used to grab attention quickly.
Question: starts with a question to involve the reader right away.
What exactly the reader should be asked to do at the end.
Ask for a comment: invite the reader to leave an opinion or reply.
How many emojis should be used in the text.
How many hashtags should be added at the end.
How strongly the text should push toward a sale.
What evidence to add so the text feels more convincing.
You can choose multiple options.
How to encourage the audience to interact with the post.
Question to the reader: the post asks a question to spark replies.
The overall personality of how the brand speaks to its audience.
Friendly expert: knows the topic well but speaks simply.
Additional sections
Named blocks for rules, constraints, examples, and any context that does not fit into the standard fields.
Named blocks for rules, constraints, examples, and any context that does not fit into the standard fields.
Sections left: 10
Use this area for brand rules, negative examples, legal constraints, or formatting instructions.
Humanization controls
Reduces recognizable AI markers, makes the tone feel more natural, and lets you explicitly exclude common AI phrases.
Reduces recognizable AI markers, makes the tone feel more natural, and lets you explicitly exclude common AI phrases.
Choose the phrases and constructions the model should avoid.
List additional phrases or constructions to avoid. Use one phrase per line.
Review and generate
Use the generated JSON prompt right away: copy it, download it, and drop it into any AI workflow or internal tool.
Result
Generated JSON prompt
Use the generated JSON prompt right away: copy it, download it, and drop it into any AI workflow or internal tool.
Social
SMM
The SMM builder is designed for teams that want structured JSON prompts for social media without rewriting the same instruction block again and again.
The page combines universal writing controls with SMM-specific settings such as platform, post goal, structure, hook, emojis, hashtags, engagement triggers, and brand voice.
This setup is especially useful for marketers, producers, editors, and brand owners who need consistent output quality across multiple channels.
FAQ
FAQ
How is the SMM builder different from a regular prompt template?
It assembles JSON on the server, separates common and SMM-specific fields, and lets you change the content strategy without editing the structure manually.
Can I use it for Reels and Telegram content?
Yes. The category includes platform and content-type options that work for feed posts, short-form video captions, and Telegram posts.
Does the SMM builder support custom values?
Yes. Every major selectable field supports a Custom value mode so you can type your own instruction.
Can I make the generated copy feel less AI-written?
Yes. The common humanization block lets you avoid long dashes, exclude typical AI phrases, and define your own phrases to avoid.