I built over 3-4 weeks to solve my own pain point, picking up rust along the way. It has a browser version that works online/offline and cross platform desktop as well. It was a pretty intense sprint.
About TreeSnap, well it simplifies this preparation of sharing code with Chat LLMs. In my use case, making o3 via ChatGPT a lot more cost-effective to use for these tasks compared to Cursor:
How it works:
Select files/folders → generates a single text bundle + ASCII tree
Instantly shows per-file and total token counts (list & treemap views)
Quickly trim unnecessary files or folders to match context limits
Fully offline and private (Tauri desktop app)
Also available as a free browser app—offline and private
Built in ~3 weeks, I use it daily to aid my workflow and reduce my expenses by minimizing the need to rely on Cursor's pricier model usage. I still rely on Cursor with grok 3 or gemini pro 2.5 pro mostly without getting crazy with usage fees. Feedback welcome—especially UX, performance, and pricing.
About TreeSnap, well it simplifies this preparation of sharing code with Chat LLMs. In my use case, making o3 via ChatGPT a lot more cost-effective to use for these tasks compared to Cursor:
How it works:
Select files/folders → generates a single text bundle + ASCII tree
Instantly shows per-file and total token counts (list & treemap views)
Quickly trim unnecessary files or folders to match context limits
Fully offline and private (Tauri desktop app)
Also available as a free browser app—offline and private
Built in ~3 weeks, I use it daily to aid my workflow and reduce my expenses by minimizing the need to rely on Cursor's pricier model usage. I still rely on Cursor with grok 3 or gemini pro 2.5 pro mostly without getting crazy with usage fees. Feedback welcome—especially UX, performance, and pricing.