# WhatsApp Parser MCP > The WhatsApp Chat Export Parser MCP takes messy chat logs exported as `.txt` files and instantly converts them into clean, structured JSON data. Your AI agent can then search years of conversation history, pull specific addresses, or summarize long threads without ever uploading your chats to the cloud. ## Overview - **Category:** talk-to-me - **Price:** Free - **Tags:** chat-export, regex-parsing, data-extraction, message-history, search-optimization, text-analysis ## Description Ever need an address your landlord sent you three months ago in a massive WhatsApp chat? Scrolling through thousands of messages is a nightmare. This MCP solves that common pain point. When you export any chat via WhatsApp's built-in feature, you get a raw `.txt` file—a mess to parse by hand. This tool reads the entire log locally, extracting every message with its exact timestamp, who sent it, and what was said. It handles both US and EU date formats automatically. The resulting structured JSON lets your AI agent search, filter, and summarize your whole conversation history instantly, all while keeping your chats private on your machine. Because this process runs within Vinkius's isolated sandbox, you keep absolute control of your data; nothing leaves your local environment during parsing. It’s a massive time-saver for anyone drowning in chat logs. ## Tools ### parse_whatsapp_chat Reads an exported WhatsApp chat text file offline, extracting every message, sender, timestamp, and participant count into structured JSON. ## Prompt Examples **Prompt:** ``` Parse my WhatsApp chat with Maria and find the message where she sent me an address. ``` **Response:** ``` Found it! On 03/12/2024 at 14:23, Maria sent: 'The address is Rua Augusta 42, 3rd floor, Lisbon.' ``` **Prompt:** ``` Summarize my conversation with the landlord from last month. ``` **Response:** ``` Last month, you discussed 3 topics: 1) Rent payment date change (agreed to the 5th), 2) Plumbing issue (he'll send someone Friday), 3) Lease renewal (pending his response). ``` **Prompt:** ``` Who sends the most messages in my group chat? Show me the statistics. ``` **Response:** ``` Participant stats: João (1,204 messages), Ana (892), Pedro (456). João sends 42% of all messages. ``` ## Capabilities ### Locate specific messages The agent searches the entire conversation history to pull out exact mentions, like finding one specific receipt or address. ### Summarize long threads It processes massive chat logs and delivers concise summaries of topics discussed over weeks or months. ### Extract metadata and stats The MCP calculates participant statistics, showing exactly who sent the most messages in a group chat. ## Use Cases ### Finding a lost address The agent needs to find the specific street address Maria mentioned months ago. You run parse_whatsapp_chat on the exported chat and ask the agent, 'Where did she say I live?' The result immediately gives you the exact date and message content. ### Summarizing lease discussions The landlord sent a 50-page chat history. You feed it to your agent and ask for a summary of 'renewal terms'. The MCP processes the data, and the agent returns a bulleted list covering all agreements. ### Determining group activity You want to know who dominated a large team chat. You use parse_whatsapp_chat on the logs, and your agent provides a clear breakdown: 'João sent 42% of messages.' The statistics are immediately available. ### Extracting decision points You need to pull every time payment dates were discussed. You run parse_whatsapp_chat, and the agent filters all records by topic, giving you a chronological list of agreement changes. ## Benefits - Never manually comb through chat logs again. The parse_whatsapp_chat tool converts unstructured text into searchable JSON, so your agent can find specific details in seconds. - Keep all data private because the MCP runs locally. Your sensitive conversations never leave your machine or get uploaded to a cloud service. - It handles date formatting automatically. Whether the chat used US (1/15/23) or EU (15/01/2023) formats, it parses them correctly every time. - The agent can summarize huge volumes of text. You get high-level insights—like key decisions made or recurring topics—without reading thousands of messages. - You don't have to worry about context limits. The tool manages massive files by focusing on core data points, preventing your AI client from crashing due to chat log size. ## How It Works The bottom line is you take a useless wall of text and turn it into a structured database your AI can actually query. 1. First, you export your desired WhatsApp conversation using the built-in 'Export Chat' feature to get a raw `.txt` file. 2. You feed that local `.txt` file into the MCP. The tool then runs dual-locale regex parsing on the text to identify and structure every message. 3. The agent receives clean, searchable JSON data containing timestamps, senders, and content, ready for immediate analysis. ## Frequently Asked Questions **Are my private chats sent to the cloud?** Never. The parsing is 100% local. Only the structured text representation is sent to the AI chat context during your session. **How do I export a WhatsApp chat?** Open the chat in WhatsApp, tap the three dots > More > Export Chat > Without Media. Save the .txt file to your computer. **What if the chat has 50,000 messages?** The engine uses a token-safe strategy: it sends only the first 100 and last 50 messages, plus full participant stats. Ask the AI to filter by sender or date for specific results. **Does `parse_whatsapp_chat` handle different global date formats?** Yes, it automatically handles multiple international date formats. The tool uses dual-locale regex parsing so you don't have to worry about US vs. European date conventions when exporting chats. **How does `parse_whatsapp_chat` process messages containing emojis or foreign languages?** It processes them fully because the parsing happens entirely local. The engine captures raw text content, including complex Unicode characters, ensuring no language barrier affects data extraction. **Can `parse_whatsapp_chat` process multiple chat files at once?** Currently, it processes one exported .txt file per call. If you have a batch of chats, you'll need to chain calls or write a simple script wrapper around the MCP tool. **What metadata does `parse_whatsapp_chat` capture for each entry?** For every message, it captures three key pieces of information: the exact timestamp, who sent it (sender), and the full content. This structure allows your agent to perform highly specific searches. **Is the parsing done on a secure or private environment when using `parse_whatsapp_chat`?** Yes, the entire process runs in a completely air-gapped sandbox. Your chats are parsed locally and never uploaded to any cloud service.