# get_annotations The get_annotations MCP tool lists every annotation in a document as JSON — id, type, message, page, bounding box, author, and replies — so an AI agent can review or edit them. `get_annotations` lists every annotation in a document as JSON. It is the **first call in almost every workflow**: the `id` values it returns are what `update_annotation`, `remove_annotations`, and `add_reply` need. Example prompt: *"What comments are in contract.pdf?"* **Tool description (as the AI agent sees it):** > Lists all annotations in a document as JSON. Supports PDF, DOCX, XLSX, PPTX, and 50+ more document and image formats. Each entry includes 'id' (use this with remove_annotations / update_annotation / add_reply), 'type', 'message', 'page' (1-based), bounding box, user, and replies. Call this tool first whenever the user asks to list, show, inspect, or review annotations or comments. Do NOT pre-check whether files exist — pass the filename the user provided directly. Returns a JSON object with `found` (count) and `annotations` (array). On failure, the response text starts with 'Annotation lookup failed for' followed by the underlying exception type, message, and inner-exception chain. ## Parameters | Name | Type | Required | Description | |---|---|---|---| | `file` | object | yes | The document — see [FileInput shape]({{< ref "annotation/mcp/tools-reference/_index.md#the-fileinput-shape" >}}) | | `password` | string | no | Password for protected documents | ## Example call ```json { "name": "get_annotations", "arguments": { "file": { "filePath": "contract.pdf" } } } ``` ## Result A JSON array, one entry per annotation: `id` (pass this to the editing tools), `type`, `message`, `page` (1-based), the bounding box, the `user` who made it, and any `replies`. That structure is what lets an agent answer *"which comments are still unanswered?"* or *"what did Maria flag on page 3?"* without opening the document — it filters the array it already holds. On failure the text starts with `Annotation listing failed for`, followed by the exception type and message. ## Example prompts * *"What comments are in contract.pdf?"* * *"List every annotation with its author and page."* * *"Which annotations have no replies yet?"* * *"Summarize the review notes on page 3."* See it used end-to-end: [Extract and report on annotations]({{< ref "annotation/mcp/use-cases/extract-and-report-annotations.md" >}}).