Flow Vault is a local-first archival viewer for Power Automate. Drop a solution export; get a searchable, visual, read-only archive that lives in your browser — no account, no cloud, no compliance conversation.
Flow Vault fixes exactly that. Export your Power Automate solutions the normal way, drop them in once, and they're yours to reference forever — every trigger, every branch, every expression, preserved verbatim in a browser tab that works offline.
It's not a runtime. It never executes a flow, never connects to a tenant, never touches Microsoft. It's an archive and a reference — the kind of tool that stays useful for years, because it doesn't depend on anything that goes away.
Every feature earned its place by answering a specific question a consultant asked. Nothing is here for the demo.
Loose .json files or entire solution zips — Flow Vault walks the Workflows folder and pulls out every workflow definition automatically.
"Which flow used formatDateTime?" Answered in milliseconds.
Batch-export or delete whole folders at once.
Parallel branches side by side, nested loops in containers, Switch cases as lanes — everything correctly indented.
Export as standalone HTML — no CDN, no dependencies. Double-click in any browser. Works offline. Works ten years from now.
Your archive moves between machines.
No server touches your flow data — you can verify this in your browser's network inspector while you use the app.
The left sidebar is where everything you have ever uploaded lives. Organize flows into folders you drag to reorder and double-click to rename. Tag by client, project, year — anything that helps you find things later.
At the top: a search box that types across every string in every flow you have ever archived. Below the folders: Backup and Restore — the two buttons that make your library portable across machines.
The middle panel is the visual rendering. Triggers at the top, actions cascading down, branches splitting on conditions, parallel siblings shown side by side, and loop bodies nested inside their containers.
Above the canvas: Expand all and Collapse all, plus a canvas-wide search, pan-and-drag hand tool, zoom, and fullscreen. At the top-right: a persistent Archived · Read-only badge — a small but deliberate reminder that you're looking at a snapshot, not a live flow.
Click any node on the canvas and its full details appear on the right: the type, the connector, every input parameter with expressions highlighted, and the raw JSON as a fallback.
Every value has a copy button. Pull an expression out of a flow you built years ago; paste it straight into a new one you're building today. That's the whole point.
No account system, no server-side data store, no telemetry, no analytics. The application is a static bundle of HTML, CSS, and JavaScript — served from a static host, and running entirely inside your browser tab.
I built Flow Vault because I lost too many flows.
Every engagement ended the same way — a client offboards you, the tenant access disappears, and years of work you spent late nights on goes behind a login you no longer have. Not stolen, exactly. Just gone. And every future project starts a little more from scratch than it needs to.
This tool exists so that doesn't happen again. Everything you archive here stays on your device. No account. No cloud. If I disappear tomorrow, your library still works — because there is no me on the other end of it. Just your browser, your machine, your flows.
That's the whole promise. That's the whole product.
No. Flow Vault is entirely client-side. Files you drop into it are read, parsed, and stored in your browser's local IndexedDB. There is no server on the other end. Open your browser's network inspector while using Flow Vault and you'll see it makes no outbound requests apart from loading its own static assets on first visit.
Yes. The current version is free. Paid tiers with optional cloud sync and shared agency libraries may exist later; the local-only experience stays free.
No. Flow Vault is not affiliated with, endorsed by, or connected to Microsoft. It reads the exported JSON that Power Automate itself produces, but it never signs in to a tenant or connects to any Microsoft service.
Two: single Power Automate workflow definitions as .json, or a whole exported solution as .zip. When you drop in a zip, Flow Vault walks the /Workflows/ folder inside it and pulls out every definition it finds. Up to fifty files per batch.
After the first visit, mostly yes. Modern browsers cache the app's static bundle so it opens without a connection. New uploads and searches happen in-browser and need no network access. Only the first load requires internet to fetch the app.
In your browser's IndexedDB, scoped to the origin Flow Vault runs on. IndexedDB is per-browser and per-user profile — not shared between Chrome and Firefox on the same machine, and not shared between two different user profiles in the same Chrome. It persists across sessions and restarts.
Clearing the cache alone is safe — it re-downloads the app but preserves your data. What you should not do without a backup is "Clear site data" or "Delete all cookies and site data," which includes IndexedDB. Export a backup .zip first and you can restore in seconds.
Whatever your browser gives IndexedDB — usually several gigabytes per origin. Flow definitions are small (tens to hundreds of KB each), so libraries of a few thousand flows are entirely reasonable. Extremely large libraries (10,000+) start to feel slow at search-index-rebuild time; the library UI stays responsive.
Yes — export a backup .zip from the current device, copy it to the other machine (or a cloud drive), and use Restore. Cloud sync as a first-class feature isn't in v1; the backup workflow is the portable-library story for now.
No, deliberately. Flow Vault is a viewer and archive. It never executes a flow, never triggers a connector, never touches a Microsoft tenant. To run a flow you'd import its .json back into a live Power Automate environment through the normal solution-import path.
Modern Chromium (Chrome, Edge, Brave, Arc), Firefox, and Safari — current or previous major version. IndexedDB, File API, and the CSS features used are stable in all of them. Internet Explorer is not supported.
No. Files are read, not written. The original .json or .zip on your disk is untouched. What's stored in IndexedDB is a copy of the parsed definition.
Yes — the search box at the top of the library sidebar searches across every string in every flow: names, tags, connector operation IDs, expression bodies, JSON comments. Answers questions like "which flow uses formatDateTime?" or "which one connects to the sales SharePoint list?"
Because it's self-contained: every style is inlined, flow data is embedded as a JS literal, the rendering logic is a small vanilla-JS mini-viewer stitched into the same document. No CDN links, no fonts to fetch. Double-click and the browser has everything.
Yes. Actions sharing the same runAfter dependency set are grouped into a level and rendered as horizontal lanes inside a dashed container labelled "Parallel" — matching Power Automate's designer.
Yes — use HTML export. You get a single .html file anyone can open in any browser without installing Flow Vault. Ideal handover artifact: read-only, self-contained, no login needed, readable years from now.
None from Flow Vault itself. Only client-side storage the app uses is IndexedDB (for your library) and small preference values in localStorage (theme, sidebar layout). No third-party analytics scripts embedded.
The current release is proprietary. Architectural choices (client-side processing, self-contained HTML export) make the trust story auditable in browser dev tools regardless. Open-sourcing is on the table for a future release; the feedback dialog is the channel to weigh in.
It renders on mobile browsers, but the interface is designed for laptop or desktop screens — flow canvases are best read at width. Mobile is a viable quick-reference mode, not the primary experience.
The Feedback button in the top bar opens the feedback dialog. Attach a screenshot for UI bugs; attach the flow's .json for rendering bugs. Feature requests use the same channel — describe the workflow you're trying to accomplish and why the app makes it hard.