render_page Leave feedback

render_page renders one page of a document as a PNG image. The image goes straight back to the agent, so a vision-capable model can look at the page itself, and a copy is saved to storage as <name>_page<N>.png. PDF, Word, Excel, PowerPoint, OpenDocument, HTML, email, and 170+ more formats are supported. Example prompt: “Show me page 3 of the quarterly deck.”

Tool description (as the AI agent sees it):

Renders a single document page as a PNG image and returns it inline (plus saves a copy to storage). Supports PDF, DOCX, XLSX, PPTX, ODT, RTF, HTML, EML, MSG, and 170+ more document formats. Call this tool immediately whenever the user asks to render, preview, or get a page image from a document. Do NOT pre-check whether files exist — just pass the filename the user provided. The tool resolves files from storage and returns an error with available files if a name is not found. Returns a CallToolResult containing both a TextContentBlock (saved file path under <source-stem>_page<N>.png) and an ImageContentBlock (the PNG bytes inline as image/png). On failure, IsError is set and the response text starts with ‘Rendering failed for’ followed by the underlying exception type, message, and inner-exception chain.

{
  "name": "render_page",
  "arguments": {
    "file": {
      "filePath": "report.pdf"
    },
    "page": 1
  }
}

Two content blocks:

• Text — Page 1 of 'report.pdf' Saved to: <output folder>/report_page1.png. In evaluation mode it starts with [Evaluation mode] Output may include watermarks.
• Image — the PNG bytes inline as image/png, which the agent can read directly.

On failure isError is set and the text starts with Rendering failed for, followed by the exception type and message.

• “Show me page 1 of report.pdf.”
• “Render the title slide of quarterly.pptx.”
• “What does the chart on page 4 of the annual report show?”
• “Render page 2 of the protected contract; the password is in my notes.”

See it used end-to-end: Preview documents with AI agents.