Diagrams that update themselves,from the Slack thread that made them.

No. Arialine writes the Mermaid for you from plain-language descriptions. You can paste Mermaid, attach a .mmd file, or export the source at any time — but you never have to touch it to use Arialine.

Diagram renders and canvases live in your Slack workspace. Arialine stores the minimum operational metadata needed for versions, decisions, access and billing, with encrypted off-site backups. The account page shows installed workspaces and subscription status.

A standard Slack app: bot user, one slash command (/arialine — settings, brainstorm and board controls), message shortcuts ('Into diagram', 'Render as canvas'), and a global shortcut ('New diagram'). It only joins channels you invite it to, and every default is the least-noisy one.

The minimum needed to post files (HTML/PNG), create canvases, read messages in channels it's invited to, and listen for message shortcuts. Full scope list is shown at install time and available in the admin docs.

Slack shows the PNG on mobile. To interact with the HTML (pan/zoom, decision badges), tap the file, download it, and open in your phone's browser. Everything else — versions, decisions, undo, rewind, /brainstorm — works from the thread on any device.

Yes. Arialine exposes a Model Context Protocol (MCP) server, so Slack's built-in assistant can create, edit, branch, and query boards on your behalf. Every AI-driven change is logged to the same decision ledger and shows who asked.

Yes. In review mode every proposed change becomes a card with Accept / Reject, and nothing modifies the diagram until an admin approves. Mode is set per channel — you can run a strict channel next to a fast-drafting one.

History is additive. Every change is v2, v3, v4… The anchor message always shows the latest render; every superseded version keeps its exact .png + .html on its thread reply. Undo and Rewind are themselves tracked as new versions, so audits always match reality.

A branch forks a version into a brand-new board with its own thread and canvas ledger, and a link back to the source. Use it to explore an alternative (e.g. 'what if we replace the retry queue with idempotent charge?') without touching the original.

/brainstorm picks a structured method (SCAMPER, Oblique Strategies, premortem, etc.) and generates 2–3 alternative takes on the current diagram. Each is a rendered preview with two buttons: Adopt in place (as a new version) or Branch to new board.

Yes. Invite the bot to the private channel, or start a DM. In DMs the ledger canvas is shared only with the DM participants.

No. Message content is used only to generate the requested render and is not used for model training. The MCP server exposes tool calls only — Slack AI decides when to call them; we never send channel history to a third-party model outside of that request.

Delete the anchor message to remove a board (its files and canvas go with it). To roll back content, use Rewind on any prior version — it becomes the new latest version, but the entire history is preserved for audit.

Yes. Arialine can be installed org-wide, and each workspace inside the org is its own set of boards with its own subscription. Audit exports of the decision ledger and a DPA are available on the Enterprise plan.

Free forever, with every feature and a cap of three boards — no member limit. Team is one flat price per workspace per month for unlimited boards, however many people use it; see the pricing page for the current amount. Enterprise is custom for Grid orgs needing invoicing, audit exports or an SLA.