Classification
Structured-data syntax stripping (value extraction)
Origin
API responses, config files, log payloads, database exports
Target
Anywhere you need prose, not data — docs, messages, notes, spreadsheets

The problem

You copy a response out of an API explorer, a browser’s network tab, or a config file, because you want the text inside it. What you get is a wall of structure: every value wrapped in quotes, every value labelled with a quoted key, and the whole thing nested inside braces and brackets and separated by commas. To reuse the words you have to manually delete all of it — a slow, error-prone edit you do the same way every time.

The Extract JSON Text action does that edit for you, the same way every time, locally on your device.

What the action strips — and what it keeps

The behaviour is deliberately simple and predictable. Given valid JSON, the extractor walks the data in order and outputs every value it finds, applying these rules:

  • Structural braces { } and brackets [ ] are removed. The object and array containers disappear; only their contents remain.
  • Keys and the quotes around them are removed. A pair like "title": "Wireless Mouse" contributes only its value.
  • Quotes around string values are removed, leaving the bare text.
  • Colons and commas — including trailing commas — are removed.
  • Each remaining value is placed on its own line, in the order it appears in the document. Numbers, true/false, and null are passed through as their plain-text form.

A worked example

Here is a small product object on the left, exactly as you might copy it from an API, and the clean text the action returns on the right.

Before · raw JSON from an API
{
  "title": "Wireless Mouse",
  "price": 24.99,
  "inStock": true,
  "tags": ["office", "ergonomic"],
  "vendor": { "name": "Acme", "rating": 4.5 }
}
syntax around every value
After · Extract JSON Text
Wireless Mouse
24.99
true
office
ergonomic
Acme
4.5
just the values, one per line

The extractor reads the structure depth-first, so nested values (the tags array, the vendor object) flatten into the same clean list in the order they appear. No braces, no keys, no quotes, no commas.

How it behaves on the edge cases

Predictability is the whole point, so the action commits to clear rules at the boundaries:

  • Nested objects and arrays are flattened in document order — their values join the same single list rather than being dropped.
  • Numbers, booleans, and null are kept as their literal text (24.99, true, null), since those are real data values.
  • Malformed or partial JSON never throws an error. If the input won’t parse cleanly, the action falls back to a best-effort pass that still removes braces, brackets, "key": labels, surrounding quotes, and trailing commas — so you always get usable text back instead of a failure.

Because the entire operation runs in your browser, the data in the blob — which is often exactly the kind of thing you don’t want to paste into a random web tool — never leaves your device.

The clean fix

When you need the words out of a JSON payload, paste it into cleanplaintext.com and press Extract JSON Text (or key 7). The syntax is stripped locally and you’re left with a clean column of values, ready to drop into a document, a message, a spreadsheet column, or your notes.