Show HN: wavecat – a fully local personal agent that watches your screen
wavecat is a fully local personal agent that monitors a user's screen activity to understand their needs and goals, with all data processing kept on-device for privacy. It highlights capabilities of local agent systems that cloud-based systems cannot offer, and requires a powerful Apple Silicon Mac or dedicated GPU to run.
Background
- The product is a "fully local personal agent" that continuously monitors your computer screen (presumably via screen capture or accessibility APIs) to infer your tasks and goals, then acts on them — all processed on-device, without sending data to the cloud.<br>- "Local" means no data leaves your machine, which contrasts with cloud-based AI agents (e.g., ChatGPT's operator mode, Microsoft Copilot) that send your activity data to remote servers for processing.<br>- "Apple Silicon Mac" or "nice GPU" is required because running large language models (LLMs) and vision models locally demands significant hardware — most consumer laptops can't handle this.<br>- The project appears to be a new open-source or indie tool (posted on Hacker News as "Show HN"), not a major company product. It targets developers and power users who value privacy and have capable hardware.<br>- The core pitch: local AI agents can offer capabilities (real-time, persistent screen understanding, no latency, complete privacy) that cloud-dependent systems cannot.