All answers
How can I automatically add text descriptions to a screen recording process?
March 6, 2026·1 min read·Screen Recording to Documentation
Glyde and Scribe automatically add text descriptions to each step of a recorded process using AI. The tool detects what you clicked, identifies the element's label, and generates a description like "Click the Submit button in the Order Details section." No manual writing needed — the descriptions appear alongside the screenshots in the generated guide.
How do auto-generated descriptions work?
| Technology | What It Does | Description Example |
|---|---|---|
| Click detection | Identifies the exact element clicked | Knows it is a button labeled "Submit" |
| Context awareness | Recognizes the page section | "in the Order Details section" |
| Action identification | Detects click, type, or navigate | "Click" vs "Enter" vs "Navigate to" |
| AI language generation | Creates natural step descriptions | "Click the 'Submit' button to confirm the order" |
When do auto-descriptions need editing?
| Scenario | Auto-Description | Edited Version |
|---|---|---|
| Clear button label | "Click the 'Save Changes' button" | No edit needed |
| Generic icon | "Click the icon" | "Click the gear icon to open Settings" |
| Data entry | "Type in the field" | "Enter the customer's email address" |
| Multiple similar elements | "Click the dropdown" | "Click the 'Status' dropdown (not the 'Priority' dropdown)" |
| Business context needed | "Click 'Approve'" | "Click 'Approve' — this triggers the fulfillment workflow" |
Most auto-descriptions are accurate for actions with clear UI labels. Editing is mainly needed for adding business context and disambiguating similar elements. A 15-step guide typically needs 3-5 minor description edits.
This answer is part of our guide to screen recording to documentation.