Eidetic Works Pro is live: persistent memory for your AI agents, $29/mo

Eidetic Works launched a Pro tier for its AI agent memory platform, offering persistent context across sessions, tools, and machines via the nucleus_ask MCP tool. The service syncs to the user's own Cloudflare R2 bucket and starts at $29 per month. The platform is designed for developers using Claude Code, Cursor, or Windsurf who need to avoid re-explaining decisions across sessions.

I just shipped Pro tier for Eidetic Works https://eidetic.works . Here's what's in it, who I built it for, and why I'm shipping three tiers $29 / $99 / $299 instead of a single plan. You're running Claude Code. You have two or three sessions open across different parts of the codebase. Session 2 doesn't know what Session 1 decided. You end up re-explaining the same architecture decisions to the same model you already talked to yesterday. This is the context-drift problem. It doesn't come from the model being bad. It comes from each session starting with no memory of what came before. nucleus ask — the MCP tool at the center of Pro — lets any agent call into your shared engram store and pull context that persists across sessions, across tools, and across machines. Managed R2 sync. Your engrams live in SQLite on your machine. Pro backs them to a Cloudflare R2 bucket with zero config. Your keys, your bucket — I'm not storing your AI history on my infrastructure. nucleus ask recall. The MCP primitive your agents use to query the shared store. One call, structured results. Wire it into your MCP config once and every session gets recall. Customer dashboard. Sync health, engram counts, last-seen surfaces. Not glamorous. Just the visibility you need to confirm it's working. Email support. Direct line. Not a help-center chatbot. 60-second restore. New machine. One command. Full engram history back, including decisions from six months ago that your codebase assumes you remember. All three live at eidetic.works/pricing https://eidetic.works/pricing right now. Stripe checkout, monthly billing, cancel anytime. Devs running Claude Code seriously. If you use it daily, you're accumulating context that the next session can't access. That's the gap Pro closes. Cursor and Windsurf power users who want recall that survives a project switch or a context-window flush. Small teams where two people use different AI tools but need shared context. One person on Claude Code, one on Cursor — nucleus ask lets both pull from the same store. nucleus-mcp 1.13.2 https://pypi.org/project/nucleus-mcp/1.13.2/ is live on PyPI right now. nucleus ask, recall activity, and recent engrams are all in 1.13.2. Install, point at your daemon, wire your MCP config. The OAuth arbitrage pattern is running in production at oauth.nucleusos.dev. If you haven't seen it — I wrote about the shim pattern here https://dev.to/nucleusos/running-claude-code-at-zero-per-token-cost-the-max-plan-oauth-shim-pattern-501a . Anthropic Max plan OAuth as a shared API surface, zero per-token cost for qualified setups. It's the same infra that powers Pro's daemon coordination. Pro is a billing tier on top of substrate that's already running. I'm not shipping a waitlist — I'm shipping a checkout. This is the part most "AI memory" tools skip. Your agent coordination context should not live inside any single vendor's cloud. Lock-in on AI tooling is bad enough. Lock-in on the memory layer that coordinates your AI tooling is worse. Pro's sync is your Cloudflare R2 bucket, your API keys. I hold none of it. The engram format is open SQLite. You can export everything as NDJSON any time via GET /export . The Pro subscription pays for managed worker infra and support, not for access to your own data. This is not a differentiator claim. It's a design constraint I'm not willing to drop for growth. No hosted daemon — your machine runs the daemon. I don't want to be in the business of hosting compute for something designed to be local-first. No cross-org engram sharing — by design. Governance first. I'll add surface area as real usage patterns surface. Early adopters will shape the roadmap. eidetic.works/pricing https://eidetic.works/pricing — checkout for any of the three tiers. If you're already running nucleus-mcp, upgrade to 1.13.2 from PyPI. nucleus ask is there now. The Pro tier unlocks managed sync and the dashboard on top. If you're new: install the daemon pip install nucleus-mcp , run eideticd -init , configure your MCP client, then upgrade to Pro once you've confirmed recall is working locally. Don't buy Pro until you've seen local recall work first — that's the foundation. Building this in public. Numbers, failures, and what breaks first — all of it goes on @nucleusos https://x.com/nucleusos .