A private Context OS that captures your screen and audio, distilling them into a structured knowledge graph — accessible from MCP, a web UI, an invisible HUD overlay, and shareable peer-to-peer between users. MIT-licensed.

Mintlify — Knowledge infrastructure for AI agents

Knowledge infrastructure for AI agents

Hey Product Hunt, I'm Igor. I built Sinain because feeding context into AI agents became more time-consuming than the work itself. My fleet of agents grew, the inbound signal grew faster, and I kept losing 5 minutes at the start of every coding session re-pasting old decisions, screenshots, and notes into the prompt. So together with my team, we built a tool that gathers it live from your screen and audio, distills it into a private knowledge graph, and serves it to any MCP-compatible agent. Worth knowing: Privacy-first: Default mode uses OpenRouter (token usage tracked). Paranoid mode runs entirely on-device with Ollama + whisper.cpp — zero network at runtime. The HUD overlay is invisible to screen capture, so it never appears in recordings or screen shares. Benchmark: 82.8% IPR on LongMemEval (ICLR 2025) Sharing: Knowledge sharing between users is peer-to-peer over WebRTC. The recipient's machine pulls from yours directly; their agents inherit the context through the same MCP. For now: macOS only (Windows on the roadmap). MIT-licensed. `npx @geravant/sinain@latest start` Happy to answer anything in the comments.

Ambient intelligence is the next frontier. How do you handle the privacy aspect of an agent that has 'eyes and ears'? Is all the processing done locally on-device?

Mintlify — Knowledge infrastructure for AI agents

Knowledge infrastructure for AI agents

Ambient intelligence is the next frontier. How do you handle the privacy aspect of an agent that has 'eyes and ears'? Is all the processing done locally on-device?

Sinain — The Ambient Intelligence

Eyes and ears for your AI agents and your teammates.

Eyes and ears for your AI agents and your teammates.