search_text_signatures Leave feedback

search_text_signatures finds embedded text signatures — stamps, labels, and native text annotations applied as signatures. It does not search the body text of the document; it looks for signature objects. Example prompt: “Is there an APPROVED stamp on this?”

Tool description (as the AI agent sees it):

Searches a document for embedded text signatures (stamps, labels, native text annotations) and returns each signature’s text content, page, position, and implementation type. Supports PDF, DOCX, XLSX, PPTX, and 30+ more document formats. Optionally filters to signatures whose text contains a specific string. Note: this searches for signature objects — to search for arbitrary text inside document content use a text-extraction tool instead. Do NOT pre-check whether the file exists — pass the filename the user provided directly. Returns a JSON object with found (count) and signatures (array with page, text, implementation, left, top, width, height per signature). On failure, the response text starts with ‘Text signature search failed for’ followed by the underlying exception type, message, and inner-exception chain.

{
  "name": "search_text_signatures",
  "arguments": {
    "file": {
      "filePath": "contract_signed.pdf"
    },
    "text": "APPROVED"
  }
}

A JSON array of text signatures: content, page, position, and implementation type.

The distinction that saves confusion: a sentence typed into the document is content; a signature object placed on the page is what this tool returns. If a search comes back empty on a document that visibly says “Approved”, the text is probably body content — not a signature.

On failure the text starts with Text signature search failed for, followed by the exception type and message.

• “Is there an APPROVED stamp on this document?”
• “List the text signatures and which page each is on.”
• “Find any signature containing my name.”