Show HN: Vim-like, hyper-efficient, LLM power tool – 20-80k tokens not 350k+
TL;DR
I built a focused TUI tool for efficient parallelized tasks, early independent testers expressed 90%+ API savings and more direct results. This is accomplished by user smaller context windows, hybrid client-side compaction, tool compaction, and system nudges (to avoid polluting the base prompts).
Nauti's Take
Small teams should test this on one repeatable terminal workflow with fixed context limits, then measure token spend, output quality, latency, and traceability against their current setup. The reported 90% savings remain unverified until reproducible benchmarks use identical tasks, models, and prompts.
Summary
I built a focused TUI tool for efficient parallelized tasks, early independent testers expressed 90%+ API savings and more direct results. This is accomplished by user smaller context windows, hybrid client-side compaction, tool compaction, and system nudges (to avoid polluting the base prompts).
My Shell features: - Vim-inspired keybindings, your keys never have to leave the keyboard (some mouse control supported as well) - Custom Markdown, LaTeX, and Mermaid rendering support - On-demand subagents with LLM-directed context handoff - Thread pooling, client-side compaction, tool compaction - Supports Powershell without a degraded experience; also support VGA/VESA mode, run even when GPU drivers break!
Sources