You don't remember filenames. You remember what's inside.

Kensaku searches your entire machine the way your brain actually works: pulls text from PDFs, Word docs, screenshots, even vague searches like "that PDF on database indexing" or "screenshots from last week."

100k filenames indexed in ~1ms. Uses 12MB RAM. Fully local—nothing leaves your device. Your AI agents can query it too.

Free. Open source.

This is how search should've worked from day one.