A lightweight web-based editor for JSON Lines (JSONL) files. Built with a focus on User Experience (UX) and data cleaning, this tool allows you to visualize, inspect, and prune JSONL datasets directly in your browser.
JSONL Studio emerged during a research session with Google's AI mode. The flow transitioned naturally from searching for an existing tool into developing this custom solution. It processes data line-by-line through a recursive table interface, allowing users to explore nested structures and prune undesired properties before exporting a cleaned JSONL file.
- Drag & Drop Interface: Seamless file loading with visual feedback.
- Recursive Table View: Objects and Arrays are not just strings; they are interactive entry points. Click an icon to open a new modal layer for deeper inspection.
- Smart Data Typing:
{}Object Badges: Deep-dive into nested structures.[]Array Badges: Specialized views for arrays of objects vs. simple lists of strings/numbers.
- Granular Pruning: Use checkboxes on any level (root or nested) to exclude specific properties from the final export.
- Modal Stack Navigation: Deeply nested exploration with an intuitive "back" flow (stacked modals).
- Privacy First: 100% Client-side. Your data never leaves your browser.
- Upload: Drag your
.jsonlfile into the drop zone. - Inspect: Preview the first rows of your dataset. Click the
{}or[]icons to explore nested data. - Clean: Uncheck the "Include" boxes for any property you wish to remove. This works recursively across all lines in the file.
- Export: Click "Export Cleaned JSONL" to generate and download a new file with your filtered schema.
No installation required.
- Clone this repository:
git clone https://github.com/Dark2C/jsonl-studio
- Open
index.htmlin any modern web browser.
This project is licensed under the MIT License.