ContextPulse captures your screen, voice, and input in real time — then delivers it to AI agents via the Model Context Protocol. One daemon, 21 tools, zero cloud dependency.
They can't see your screen, hear your voice, or remember what you were doing five minutes ago. Every conversation starts from scratch. Every context switch means re-explaining everything.
Your agent has no idea what app you're in, what error is on screen, or what you're looking at right now.
Voice context vanishes the moment you stop talking. Your agent can't reference what you said two minutes ago.
Every session resets. Your agent doesn't know what you were doing before, what you copied, or what you typed.
ContextPulse runs as a single lightweight daemon in your system tray. It captures everything your AI agent needs to understand your working context.
Auto-captures your screen every 5 seconds with full OCR. Tracks clipboard changes, active windows, and app usage patterns.
Hold-to-dictate with local Whisper transcription. Self-improving vocabulary learns your jargon and auto-corrects misheard words.
Monitors typing patterns and mouse activity. Detects when you correct voice dictation errors and feeds improvements back to Voice.
ContextPulse exposes 21 tools through the Model Context Protocol. No APIs to configure, no tokens to manage. Your AI agent calls ContextPulse tools just like any other MCP server.
{
"mcpServers": {
"contextpulse-sight": {
"command": "contextpulse-sight-mcp",
"transport": "stdio"
},
"contextpulse-voice": {
"command": "contextpulse-voice-mcp",
"transport": "stdio"
},
"contextpulse-touch": {
"command": "contextpulse-touch-mcp",
"transport": "stdio"
}
}
}Runs silently in your system tray. Captures screen, voice, and input 24/7 with less than 1% CPU usage.
Search across screen, voice, clipboard, and keyboard events in a single query. Find anything you saw, said, or typed.
Temporal view of all events across modalities. Reconstruct exactly what happened, when, and in what order.
All data stays on your machine. Local SQLite database, local Whisper model, zero cloud uploads. Your context is yours.
Voice dictation improves over time. Learns your project names, technical terms, and common phrases automatically.
Touch detects when you fix voice dictation errors and feeds corrections back. Your agent gets smarter every session.
The core experience is free forever. Pro unlocks cross-modal superpowers for power users.
Everything you need to give your AI agents context awareness.
Full cross-modal awareness for professional workflows.
No spam. We'll notify you when Pro launches.
Download ContextPulse and start your first session in under a minute.