Skip to main content

Bowrain

Bowrain shares writing guidance, terms and content memory across projects. Its context graph connects content to the rules that apply to its audience, channel, language and other coordinates. Teams work through web and desktop clients backed by a shared server.

Start with The context graph for the model and an example of applying different rules to the same feature name.

Connectors import content from content platforms, design tools, repositories and local checkouts. Imported content uses the same editing, review and version history features regardless of its source.

Outdated wording

The narration uses terminology that has since been retired. The interface and the behaviour shown are unchanged.

What Bowrain adds​

Shared, versioned governance​

Projects share server-hosted voice profiles, terms and content memory. Version history records changes and approvals. Recurring corrections provide evidence for candidate rules that reviewers can promote into a profile. See Voice and corrections.

Governed review​

A run drafts and checks content, and holds back what a machine cannot decide. A person approves those items in one review session, across items and locales, and Bowrain records each decision against the workspace. The last approval for a locale starts the run that delivers it.

Collaboration​

A web editor and a desktop app connect to the same server. People work in Visual and Table views with the checks, the terms in force and the content memory in context; edits, presence, and live updates propagate to every client without a refresh. See Real-time collaboration.

Connectors​

A workspace syncs against the systems content already lives in: a content platform (WordPress, HubSpot), a design tool (Figma), a code repository (GitHub, GitLab), or a developer's checkout (kapi). Most run server-side, with no local checkout: source flows into the governed workspace, gets drafted and checked, and publishes back to where it came from. See Connectors.

Context for AI assistants​

AI assistants can retrieve workspace terms, content memory and voice guidance over MCP. Checks assess generated text against that guidance, and recorded corrections provide evidence for future rules.

Automation​

Server-side rules run flows when content arrives, when a language stops meeting a ship gate, or when a review completes. The rules apply across the workspace. See Automation.

Stewardship over time​

The content store records block history across branches and languages. Terms, corrections and approved wording remain available as project contributors change.

How it fits together​

Content sourcesCMS · design · repository · checkoutBowrain workspacegovern · automatePeople & agentsweb / desktop / MCPbring source ina connector fetch, or a pushdraft, check, gateserver automationdeliver for reviewlive-sync to web / desktopreview & approvein the sessionapproved editsback to the workspacepublish backto the source system

Content flows in from any connector, the server governs and automates, people and AI tools review, and results publish back to where they came from.

Parts of the platform​

  • Server: workspaces, authentication, the versioned content store, automation, and the connectors.
  • Connectors: content platforms, design tools, code repositories, and a developer's checkout. Most run server-side with no checkout involved; see Connectors.
  • Web app: the Context hub, the review session, and a browser-based editor with Visual and Table views.
  • Desktop app: a native cross-platform client with offline support.

What's next​

Bowrain and kapi​

Bowrain is built on kapi, the Apache-2.0 toolchain that holds the context graph for one project. kapi works on local files and needs no server or account. Bowrain adds shared context, hosted review and collaboration across projects.