Vnsh

•
156
•

Provides ephemeral, client-side encrypted sharing for AI context like logs, diffs, and images, vaporizing data within 24 hours.

Vnsh serves as a secure, ephemeral "Dropbox" designed specifically for AI workflows, eliminating the need to paste sensitive or lengthy data directly into AI prompts. It employs a Zero-Access Architecture, ensuring all AES-256-CBC encryption occurs on the client device, with decryption keys never leaving the user's browser or CLI and never reaching the server. The platform stores only blind, encrypted blobs that automatically vaporize after 24 hours, guaranteeing privacy and preventing data leaks or server compromise from exposing content. It's ideal for sharing debugging context, code diffs, large logs, screenshots, and even binaries securely and temporarily with AI assistants.

Características Principales

014 GitHub stars
02Host-blind architecture: decryption keys travel only in URL fragments, never sent to servers
03Versatile CLI: pipe logs, diffs, and upload various files from the terminal
04Automatic data vaporization: all shared data auto-destructs after 24 hours
05Native Claude Code (MCP) integration for seamless AI interaction
06Client-side AES-256-CBC encryption for ultimate privacy
07GitHub Action for automated CI/CD log sharing and debugging

Casos de Uso

01Automate the upload and sharing of CI/CD failure logs to GitHub pull requests
02Securely share sensitive logs, code diffs, or screenshots with AI models like Claude for debugging and analysis
03Temporarily share large files, configuration details, or other ephemeral context without long-term storage or privacy concerns