A1 AI engineering company GitHub
Connect the Google Slides MCP server to your AI agent Open source, MIT, on npm. Setup instructions for Claude, Cursor and Codex are in the repository; adoption questions — Telegram. GitHub repository Telegram

Google Slides: Find & replace text — MCP tool

Google Slides MCP tool: Finds and replaces every occurrence of a string across the whole deck — shapes and table cells on slides, and their speaker notes (batchUpdate replaceAllText).

Technical name: replace_text

What task it solves

I want to find and replace text across the deck.

Finds and replaces every occurrence of a string across the whole deck — shapes and table cells on slides, and their speaker notes (batchUpdate replaceAllText).

When to use it

Use this capability when you need “Find & replace text” without doing the same work manually in the Google Slides interface. It runs only when an AI client calls it.

What to provide

  • presentation_id — required. The presentation id — the long id from the URL (docs.google.com/presentation/d/<presentationId>/edit) or from create_presentation output.
  • find — required. The literal text to find (no regex).
  • replace — required. The replacement text (empty string deletes the matches).
  • match_case — optional. Case-sensitive matching (default true).
  • page_object_ids — optional. Limit the replacement to these slides/pages (default: the whole presentation).

What it returns

Returns compact JSON from the upstream API or a clear MCP tool error. The exact fields depend on the operation and are documented in the technical reference.

What changes in Google Slides

The source marks the entire “Find & replace text” call as destructive: it can permanently remove or overwrite existing content. Review the parameters and reversibility before calling it.

Example request

Find & replace text in Google Slides. Ask for any required identifiers that are missing. Show me the exact change and wait for confirmation first.

Errors and limitations

Scope it with page_object_ids to touch only specific slides. Matching is literal (no regex), case-sensitive by default (match_case=false to ignore case). Returns occurrencesChanged in the reply — 0 means nothing matched, check the exact wording with list_slides. The classic template flow: duplicate a template deck, then replace {{placeholders}}.

Access also depends on token permissions, quotas, and upstream API limits.

Related MCP tools

Technical details