llms.txt Content
<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>
vision-memory-mcp — Visual Cache & Perceptual Hashing MCP Server
</title>
<meta
name="description"
content="Persistent visual cache for LLM-driven software development. Caches screenshots using perceptual hashing, vector search, and AX trees to prevent token overhead and visual hallucination loops."
/>
<meta
name="keywords"
content="vision-memory-mcp, mcp, model context protocol, visual cache, perceptual hashing, lancedb, clip embeddings, llm vision, token optimization, agentic coding"
/>
<meta name="author" content="PuterVision LLC" />
<meta name="robots" content="index, follow" />
<meta name="theme-color" content="#0a0b10" />
<meta http-equiv="Content-Security-Policy" content="default-src 'self'; style-src 'self' 'unsafe-inline' https://fonts.googleapis.com; font-src 'self' https://fonts.gstatic.com; script-src 'self' 'unsafe-inline' https://unpkg.com; img-src 'self' data: https:;" />
<!-- Canonical URL -->
<link
rel="canonical"
href="https://visionmemorymcp.com/"
/>
<!-- JSON-LD Structured Data for Search Engines & AI Crawlers -->
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "SoftwareApplication",
"name": "Visual Memory MCP",
"operatingSystem": "Cross-platform (Linux, macOS, Windows)",
"applicationCategory": "DeveloperApplication",
"description": "Persistent visual state cache, perceptual hashing server, and visual baseline verifier for multimodal developer AI agents. Caches screenshots locally to optimize context windows and reduce vision token overhead.",
"softwareVersion": "0.9.0",
"programmingLanguage": "TypeScript",
"runtimePlatform": "Node.js >= 18",
"codeRepository": "https://github.com/putervision/vi