Sensei watches your sessions with AI assistants — then surfaces the patterns you're too close to see. Not a chatbot. Not a copilot. A patient observer.
Three sessions corrected this week in lumen-auth — all touched refresh or device flow. There's no integration-test persona for this module yet.
Sensei runs on your machine and observes your sessions with AI assistants. It logs nothing remotely; it speaks rarely; it remembers what you've actually done.
Over weeks, it begins to recognize your patterns — the idioms you gravitate toward, the workarounds you've adopted, the friction points that keep recurring. When something looks worth noticing, it tells you. The rest of the time, it stays out of the way.
Each surface answers one question and stays quiet otherwise. The pixels will keep changing; what they're for won't. So here is what each one does, why it's shaped that way, and how its flow moves.
Out of everything sensei watched overnight, Today elevates a single focal observation and keeps the rest out of sight. Most mornings it is nearly empty — that emptiness is earned, and it is the point.
Every session sensei witnessed, digested into a retrospective you didn't have to run. Not a log to scroll — a verdict, in plain language, with the trend underneath.
The patterns sensei is tracking but hasn't adopted: each with a confidence, the number of projects it's appeared in, and how long it's been forming. Sensei proposes; you dispose.
An adopted memory is a small, named lesson sensei applies to future matching sessions — with your blessing. Every one is auditable down to the moments that formed it.
An assistant is only as good as the instruments it can reach. Instruments lets you exercise each MCP tool in isolation and replay exactly what was called during a session.
AI tools are getting louder. More suggestions, more autocompletes, more interrupting. Sensei moves the other way. It speaks rarely, and only when it has something specific to say. Most days it is completely silent — and that is the feature.
The kanji throughout the app are not decoration. Each one names a phase of practice — observation, recognition, adoption, refinement. They are what we ask of the user, and what we ask of ourselves as the people who built this.
Sensei is in early preview — we're learning what works and what doesn't. It's free while we figure that out. If we move to a paid tier, early adopters and supporters get a permanent discount. No surprises.
Sensei is built by a small team. A sponsorship funds the focused hours that keep it sharp.
♥ Sponsor on GitHub