✂️ LeanPrompt — Local Prompt Optimizer

Optimize AI prompts 100% on your device. Reduce token usage, cut AI costs, and keep your data completely private — zero cloud calls, ever.

✨ Features

🦙 Ollama-Exclusive — Connects to your local Ollama server for high-quality, LLM-powered prompt optimization with full context awareness
💰 Cost Tracker — Tracks tokens saved and estimated USD savings per session
📁 File References — Type @ in the prompt box to search and reference workspace files inline
🔒 100% Private — Your prompts never leave your machine. No telemetry, no API keys, no cloud
⚡ Live Status — Automatically detects when Ollama starts or stops and updates the UI in real-time

Open the LeanPrompt panel from the Activity Bar (scissors icon)
Paste your prompt into the text box (or type @filename to reference a workspace file)
Choose Aggressiveness: 🛡 Conservative | ⚖️ Balanced | 🔥 Aggressive
Enable Safe Zone to protect code blocks, URLs, and file paths
Click Optimize Prompt
Review the Before/After comparison and token savings
Click Copy or Send to Chat to use your optimized prompt

You can quickly optimize your current selection using the keyboard:

Setting	Default	Description
`leanPrompt.compressionAggressiveness`	`aggressive`	How aggressively to compress (`conservative`, `balanced`, `aggressive`)
`leanPrompt.preserveCodeBlocks`	`true`	Protect code blocks, URLs, and file paths from compression
`leanPrompt.costPerMillionTokens`	`2.5`	Reference cost (USD) per 1M tokens for savings calculations

LeanPrompt is designed from the ground up to be 100% offline and private:

No API keys required — works entirely with local models
No network calls for inference — all optimization runs on your CPU
No telemetry — your prompts, files, and savings data never leave your machine
Your data is yours. Always.

If you find LeanPrompt useful and want to support its development, you can buy me a coffee!

Found a bug or have a feature idea? Open an issue on GitLab.