# Docparser MCP for AI Agents AI Agent Connect > Docparser MCP lets your AI agent pull structured data out of PDFs, images, and scanned docs. It handles the heavy lifting of OCR and layout parsing so your agent can focus on the data. You can manage your parsers, check extraction statuses, and pull results directly into your workflow without manual copy-pasting. ## Overview - **Category:** productivity - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_miP3tpayMODTxLaEmawpsldPcHjNW5qK9YgNrZ8w/ai-agent-connect - **Tags:** data-extraction, ocr, pdf-parsing, automated-workflows, structured-data, data-processing ## Description Stop wasting time manually copying data from messy PDFs or scanned invoices into spreadsheets. This Connector connects your AI client to Docparser, turning unstructured documents into clean, usable data instantly. When you ask your agent to pull info from a batch of files, it handles the connection, identifies the correct parser, and grabs the fields you actually need. You can check if a specific document finished processing, find out why an extraction failed, or pull the latest results across your entire account. It’s about making document data searchable and actionable without the manual friction. Instead of opening a dozen tabs to find a specific invoice number or a shipping date, your agent can just pull that info into the chat. This saves your team from the tedious find and copy loop that slows down operations. If you're looking to scale how your team handles high-volume paperwork, having this in your Vinkius catalog gives your agent the eyes it needs to see into your files and the hands to move that data where it needs to go. ## Tools ### get_docparser_account_metadata Get your account's current usage limits and general metadata. This helps you keep track of your remaining quota and account status. ### list_recent_extractions Get a chronological feed of the latest results from all your parsers. This is perfect for seeing the most recent data pulled into your system. ### quick_parser_health_audit See a high-level summary of how your parsers are performing right now. It's a fast way to check overall system health. ### get_document_extraction_results Fetch the actual data extracted from a specific document you've processed. This lets your agent pull specific fields like totals or dates directly into your chat. ### list_parsed_documents See a list of every document a specific parser has finished handling. This is useful for verifying that a batch job is complete. ### list_failed_document_extractions Identify which documents didn't make it through the parsing process. Use this to quickly find files that need to be re-uploaded or corrected. ### list_document_parsers See every parser you've configured in your Docparser account. This gives you a bird's-eye view of all your active extraction rules. ### list_documents_awaiting_parsing Check which files are currently sitting in the parsing queue. It helps you monitor the flow of documents waiting to be processed. ### search_parsed_documents Find a specific document by its filename within a particular parser. This makes it easy to locate individual files without scrolling through lists. ### get_parser_details Pull the specific settings and current status for one of your parsers. You can use this to check if a parser is active or what rules it's following. ## Prompt Examples **Prompt:** ``` Are there any issues with my parsers right now? ``` **Response:** ``` I checked your parser health. Here's the summary: - **Invoices Parser**: 98% success rate (Healthy) - **Shipping Labels**: 45% success rate (Warning) It looks like the Shipping Labels parser is struggling. Would you like me to list the failed extractions for that one? ``` **Prompt:** ``` Find the total price for the invoice 'INV-992' in the Invoices parser. ``` **Response:** ``` I've located 'INV-992' in your Invoices parser. **Extracted Data:** - **Invoice Number:** INV-992 - **Date:** 2023-10-12 - **Total Price:** $1,450.00 Would you like the full JSON output for this document? ``` **Prompt:** ``` How many documents are still waiting to be processed? ``` **Response:** ``` There are currently 14 documents sitting in your parsing queue. **Queue Status:** - **Pending:** 14 - **Active:** 2 - **Completed Today:** 128 I can list the specific documents waiting if you'd like. ``` ## Capabilities ### Pull structured data from documents Your agent can grab specific fields like totals, dates, or line items directly from a processed file. ### Check processing queue status See which files are currently waiting in the queue and which ones have finished. ### Audit parser success rates Get a high-level summary of how well your different parsers are performing. ### Search for specific files Find a specific document by its filename within a particular parser instantly. ### Retrieve account usage limits Check your current account metadata and remaining processing quotas. ## Use Cases ### Extracting totals from a batch of invoices An operations manager asks the agent to find all invoices from a specific supplier and sum the totals. The agent uses list_recent_extractions and get_document_extraction_results to do it. ### Finding a specific report by name A data analyst needs to find a specific report from last month. The agent uses search_parsed_documents to locate the file and then pulls the structured fields for a summary. ### Diagnosing a drop in data flow An automation lead notices a drop in data flow. The agent uses quick_parser_health_audit and list_failed_document_extractions to find the specific parser that's failing. ### Checking if a batch job is complete A user wants to know if their daily batch is done. The agent uses list_documents_awaiting_parsing to check the queue and list_parsed_documents to confirm completion. ## Benefits - Stop manual data entry by using get_document_extraction_results to pull fields directly into your internal apps. - Monitor your entire pipeline health with quick_parser_health_audit to catch errors before they pile up. - Save time searching for files by using search_parsed_documents to find specific records by name. - Manage your parser fleet easily with list_document_parsers to see all your active rules in one place. - Identify processing bottlenecks quickly using list_documents_awaiting_parsing to see what's stuck in the queue. ## How It Works The bottom line is that your AI agent gains the ability to treat your document library like a searchable database. 1. Add your Docparser API Key to your AI client's configuration. 2. Connect the Connector through the Vinkius platform. 3. Ask your agent to fetch data or check parser health using natural language. ## Frequently Asked Questions **Can the Docparser MCP extract data from scanned images?** Yes, it can. Since Docparser supports OCR, your AI agent can pull structured data from photos, scanned PDFs, and even low-quality images. **How does Docparser MCP help with invoice processing?** It automates the data entry part. Your agent can fetch invoice numbers, dates, and totals from your Docparser account and present them in a chat or a table. **Can I use Docparser MCP to see which documents failed?** Yes, the Connector includes a tool to list failed extractions. This helps you quickly identify which files need to be re-uploaded or fixed. **Does Docparser MCP work with my existing Docparser account?** Absolutely. You just need your Docparser API key to connect the Connector to your existing parsers and document history. **Can I search for specific files using the Docparser MCP?** Yes, you can ask your agent to find a document by its filename within a specific parser, making it much faster than browsing folders. **What happens to my data when I use Docparser MCP?** The Connector acts as a bridge. It retrieves the data you've already processed in Docparser and brings it into your AI client for you to use. **How do I get a Docparser API Key?** Log in to your Docparser account, navigate to the API section in your settings, and you can retrieve your unique API Key from there. **What types of data can be extracted?** Docparser can extract text, numbers, dates, and even complex table data from your documents based on the rules you configure in your parsers. **Can the agent show real-time processing status?** Yes, you can use the list_parsed_documents or list_documents_awaiting_parsing tools to see where your documents are in the extraction pipeline.