Actions14
Snapshot
AI-generatedSummary
Capture an accessibility tree snapshot from a web page tab, optionally scoped and filtered, to obtain a token-efficient representation of the page's structure and content.
Inputs
- Tab ID — Specify the target tab by its ID; leave empty to use the active tab.
- Filter — Choose to capture all elements or only interactive elements like buttons, links, and inputs.
- Format — Select between full or compact snapshot format; compact is more token-efficient.
- CSS Selector — Scope the snapshot to a specific part of the page using a CSS selector.
- Diff Only — Return only changes since the last snapshot instead of the full snapshot.
- Max Tokens — Limit the size of the returned snapshot approximately by number of tokens; zero means no limit.
Output shape
a JSON object representing the accessibility snapshot content
The output JSON contains the accessibility tree data filtered and formatted per the input options. If 'Diff Only' is enabled, only changes since the last snapshot are returned. Errors are returned in the JSON for individual items when 'Continue on Fail' is enabled.