FolioCradle

Atlassian apps

Bulk edits that don't break your pages.

Change hundreds of Confluence pages at once, and change nothing you didn't ask for.

Status

The first app is built and running. It is not yet published on the Atlassian Marketplace — this page will link to the listing when it is.

A foliocradle is the padded rest a conservator uses to hold a manuscript open without cracking its spine.

The problem with bulk editing

Search and replace across a wiki sounds simple, and the naive implementation is a disaster. Treat a Confluence page as text and run a regular expression through it and you will rewrite things that were never prose: the parameters inside a Table of Contents macro, the configuration of a third-party macro, the contents of a code sample. Pages come back mangled, or empty.

Walk only the visible text instead and you get the opposite failure. Every occurrence hiding in a link target or a smart-link card is silently missed, so the tool reports success while leaving half the job undone.

Both failures are common enough that the category has a reputation. FolioCradle exists because both are avoidable.

How it works

  1. 1 Scope it A space key, or a CQL query when you need something narrower.
  2. 2 Preview every change A dry run showing each occurrence, page by page, before and after — and where it lives: body text, a link target, a card URL. Nothing is written.
  3. 3 Approve the pages you want Deselect any page and it is left completely untouched.
  4. 4 Apply Each edit becomes a normal Confluence version with a descriptive message, so page history can revert it like any other change.

The preview and the apply run the identical engine, and apply re-reads each page live rather than trusting what the browser sent. The diff you approved is the change you get.

What is never touched

Macro configuration

Macro parameters are configuration, not prose. They are reproduced exactly — including for third-party macros nobody has ever heard of.

Code blocks

Opaque by default. A base-URL migration may legitimately want them, so it is a checkbox — off unless you turn it on.

Page structure

Tables keep their merged cells and column widths. Layouts, panels and nesting are preserved node for node.

Anything unmatched

A node the tool was not asked to change is copied through untouched — type, attributes, formatting, child order.

That last point is the one worth testing, so it is tested directly: the document is reduced to a skeleton with the permitted edit sites blanked out, before and after, and the two must be byte-for-byte identical. It is a much stronger claim than checking a handful of nodes, and it runs on every build.

Your content never leaves Atlassian

FolioCradle apps are built on Atlassian Forge and run inside Atlassian's own infrastructure. There is no server of ours in the path, because there is no server of ours at all.

Data sent to usNone. The app declares no external network permissions, so the Forge platform blocks outbound calls. Page content cannot reach a third party even in principle.
Data storedNone. No database, no cache, no logs of your content. Page text exists only in memory for the seconds a request takes.
Analytics or trackersNone, in the app or on this website. This page loads no fonts, scripts or images from anywhere else.
What it can reachOnly what you can. The app acts as the signed-in user, so existing page restrictions apply unchanged — it can never edit something you couldn't edit yourself.

Full detail in the privacy policy.

Apps

Bulk Find & Replace Coming soon

Find and replace across a whole space or a CQL-scoped set of pages. Rewrites body text, link targets and smart-link URLs; leaves macros, structure and code alone. Case-insensitive and whole-word matching, per-page approval, and a dry run before anything is written.

More is planned in the same family — bulk labels, link remediation, metadata operations, macro migration — all built on the same engine and the same rule about what must never change.

Questions, or something you wish these apps did? Get in touch.