StoryOS

What StoryOS is

StoryOS is an open-source work operating system built on a connected database, with AI agents that operate inside the workspace. Its site describes it as "open-source work OS with AI agents, self-hosted or cloud" and "the operating system for your business." You model clients, projects, content and deals as real connected data, look at them as boards, calendars and timelines, and then let an AI agent do part of the work.

It speaks to founders, agencies and small teams that want to own the system they run their business on: the AGPL-3.0 license, a one-command self-host and no record limits. The pages do not say where the company is based or how large it is. The site states that its own product backlog, its website's issue tracker and its content pipeline all run in a StoryOS workspace.

What you build and how

The data layer is described as "a real database under everything." You link databases with true one-to-many and many-to-many relations, roll up and look up values across them, and compute fields with formulas written in a safe expression language with about 20 functions. Rollups can aggregate a whole collection with sum, count, min, max or avg. The site sums it up as the power of Airtable and the structure of Notion, open source and driven entirely by an API.

The same records show up through 8 view types: table, board, calendar, gallery, timeline, list, feed and form. The table is the typed spreadsheet with inline edit. The board drags work through stages and keeps counts visible. The calendar puts anything with a date on its day, the gallery shows rich cards, the timeline draws bars on a zoomable axis, the list gives dense groupable rows, the feed is a review stream where you approve drafts card by card, and the form turns a database into an intake channel. Views are saved per database with filters, multi-sort, grouping and color-by. Dashboards add number tiles and charts across your data.

Automation comes from rules that run on record events or a schedule, plus one-click buttons for manual actions. Records can carry rich-text descriptions, @mentions, comments, and dedicated attachment fields for a cover, a video or a gallery, and a workspace can also hold standalone documents.

The AI layer is the headline. StoryOS ships an MCP server, so Claude and ChatGPT can read your schema, query records, and create, update, link and move work. The site's framing is that the workspace is the memory, the agent does the labor and you approve the moves that matter. Its worked example is a content pipeline that runs idea, research, brief, draft, SEO, review and schedule, with the agent moving records across the board. You connect your own Claude or ChatGPT, and the company says it never meters or marks up that AI. Its comparison page adds that 161 tools let agents build databases and automations, not just fill them.

Packs are the templates: Agency OS, Client Portal, Dev Project OS, Content Engine, Book Launch and Coaching OS are named, out of 8 packs on the site. Integrations named on the site are GitHub, Linear, Slack and YouTube, plus a full REST API and the MCP server on every plan.

For the open source category, the facts are specific. The site says "Genuinely open source AGPL-3.0, the full product," and describes self-hosting as one Docker command on your own box, with every feature and unlimited records, the same code as the hosted version. Capability is not paywalled either way. Managed hosting adds backups and a hosted MCP endpoint.

On the end-user side, viewers and commenters are always free, so a client or stakeholder can look at work and leave feedback without a paid seat. The pages do not describe publishing a public website or a branded client app from a workspace.

The evidence from the pages

Pricing is flat per workspace. The site lists $29 for Pro and $99 for Business, and says Business is $99 per workspace, not per seat. Builder seats beyond the included ones are a flat $12 extra. You start free: the site mentions a free plan, 30 days of Pro, no credit card, and a hosted workspace that starts free with the 30 days of Pro included. The pages do not list what the free plan allows after those 30 days.

The company says there are no record limits on any plan, and that viewers and commenters are free on every plan. Its comparison table, dated September 2026, sets this against per-seat pricing at Airtable ($20 to $45 per user per month, with record limits of 1,000 to 125,000 per base by tier), Notion (about $10 to $20 per user per month) and Fibery (€12 to 40 per user per month). Those competitor figures are the company's reading of public pricing pages, so check the vendors directly.

The Fibery page also says plainly when Fibery is the better choice: if you rely on its product-management flows from feedback to insights to features, need whiteboards and product discovery tools, or want EU-hosted managed SaaS with per-user pricing. It gives a 4-step move from Fibery, from CSV export to an import wizard to re-wiring relations and rebuilding automations. The pages carry no security certifications and no customer counts. Support points to docs and the GitHub repository.

Who it suits and who it does not

StoryOS suits a founder or agency that wants one database for clients, projects and content, plans to let AI agents operate it, and prefers a flat bill with free viewers. Teams that want to self-host and read the source are well served, since the AGPL-3.0 license and Docker install are part of the pitch.

A solo founder can start on the 30-day Pro trial, then choose between the hosted Pro plan at $29 and self-hosting, which the site calls free forever. An agency can give clients viewer or commenter access without paying for seats and keep the $12 builder seats for its own staff beyond those included in the plan.

The cautions are fair ones. Read what AGPL-3.0 requires before you modify and redistribute the code. The site names 4 integrations and then says "and more", so check that your own tools are covered. Teams that need deep product-management tooling may find the vendor's own comparison sends them to Fibery.