OzBrain — the brain behind every agent

Author: OzBrain (product site, no named author)

OzBrain is a hosted "shared brain" every AI agent can read and write. The pitch: one structured knowledge base — articles with links, provenance, and freshness — that Claude, ChatGPT, Cursor, Claude Code, and any connector-speaking client all draw from, so you never ferry context between tools by hand.

The core diagnosis is the context-ferrying job: copy a brief into Claude, paste it into ChatGPT, drop the same .md into Cursor, and watch the copies drift. The product's answer is a single source of truth that "puts the current article in front so every agent decides from the same file."

The distinctive claims are (1) agents maintain the brain rather than humans — "continuous maintenance as the designed behavior," with write-time size discipline so agents read only what they need; (2) a contrast with platform memory, which "keeps scraps and summaries" while OzBrain "holds the work itself"; and (3) a trust story — encrypted at rest, never trained on, per-access audit log, markdown export anytime, and delete-means-deleted. Pricing is freemium: Free (50 articles), Pro $20/mo (300), Max $99/mo (600), and a custom Company tier, all with unlimited reads/writes and brains.

It positions itself not as a memory API for developers (explicitly "not a service you code against" — the Mem0/Supermemory distinction) but as a connector you add from the Claude or ChatGPT menu via a single MCP URL.