List Studio — MVP Documentation

Stable • Feature-Complete • No Backend Required

Overview

List Studio is a browser-based tool for creating, validating, and cleaning large AI-generated lists before publishing.

It is designed for content creators, publishers, and editors who work with long lists (50–1000+ items) and need to ensure quality, uniqueness, and structure without manual cleanup.

Privacy First: No accounts. No backend. No data leaves your machine.

What This Tool Does

Builds deterministic, structured prompts for list generation
Enforces JSON output from LLMs
Validates pasted AI output
Detects exact duplicates
Detects semantically similar items
Forces human review where ambiguity exists
Produces a clean, publication-ready list
Exports the final list in multiple formats

How to Use

Open index.html in a modern browser
Fill in the Prompt Builder fields
Click "Generate Prompt"
Copy the prompt and paste it into your LLM
Generate the list
Paste the full JSON response into "LLM Output Intake"
Click "Validate & Analyze"

If duplicates are found: Select ONE item per duplicate group to keep

If similar items are found: Review and resolve as needed

Once resolved: The final list is generated automatically

Export using JSON, Markdown, HTML, or PDF buttons.

Input Requirements

Output MUST be valid JSON
Top-level structure MUST be an array
Each item should contain:
- id
- item (text)
- category (optional but recommended)

Invalid JSON will be rejected.

Example Format:

[
  {
    "id": 1,
    "item": "Freelance social media management",
    "description": "Manage Instagram/TikTok for small businesses",
    "category": "Side Hustles"
  }
]

What This Tool Does NOT Do

No automatic deletions
No AI rewriting
No hidden transformations
No cloud processing
No opinionated filtering

You are always in control.

Technical Notes

Pure HTML / CSS / JavaScript
jsPDF used for PDF export only
Designed to run locally or from any static host
No build step required
Optimized for lists up to 300 items (larger lists may experience slower similarity checks)

Intended Use Cases

Large idea lists
Side-hustle compilations
Content audits
AI-generated drafts before publishing
Editorial QA for list-based content

Status

This is a complete MVP.

Feature-complete by design
Stable
Deterministic
No planned additions required for core functionality

License & Usage

Internal tool. Distribution and usage at publisher's discretion.

License: D-BAD License

A license for developers who want to be bad.