The Monday command center for solo creators.
Today's content, this month's money, and the next thing to ship, in one private workspace. CMD-OS puts the content calendar, the clients and invoices, and an assistant that drafts from all of it on one Monday screen, with a private MCP server per workspace.
The problem
A solo creator's Monday is spread across a content calendar, a notes app, a CRM, invoicing and a scheduler. The answers to “what is due, what do I post, what got paid” live in five places, and an AI writing tool sees none of them.
Who it's for
Solo creators and independent operators who already juggle a content calendar, clients and invoices. Not agencies running courses on a platform, and not people who want a Notion template.
The product
- Core: Today view, tasks, rituals, goals, notes and a ⌘K command palette.
- CRM: pipeline, follow-up cadences, mail, proposals and invoices by link, booking pages, lead forms.
- Content: podcast, YouTube and social posts for X, LinkedIn, Instagram, Facebook, TikTok and Threads.
- AI: drafts captions, posts, week plans and research briefs from the workspace, in a saved brand voice; daily recaps and semantic search.
- Integrations: real-time GitHub activity, Google Calendar, Gmail.
- Agent access: each workspace is its own MCP server for Claude, ChatGPT, Grok, Cursor and VS Code.
Why now
AI assistants are becoming how people do their work, and MCP has become the common way to connect them to tools. Assistants need a system of record they can safely act on. CMD-OS is one: a private server URL, consent per app, tools scoped to what's switched on, and no ability to delete data or touch money.
Business model
- Three self-serve plans through Stripe Checkout and the customer portal; annual is ten months' price. No free tier.
- AI is metered in runs per period, with a $10 top-up for 80 more. First 100 Solo customers lock $19/month.
- Room to grow (not yet built): team invites to fill Studio's seats.
Why it's defensible
- One data model across clients, content, notes and code: the context AI tools need.
- Per-workspace MCP and OAuth, built in rather than bolted on.
- Tenant isolation enforced by the database, not application code alone.
Technology
- TypeScript and Next.js 15; Postgres 16 with row-level security and pgvector.
- 60,000+ lines of application code and 50+ automated test suites, including isolation tests.
- Stripe billing, encrypted credentials, and it deploys to Vercel, Fly.io or Docker.