> For the complete documentation index, see [llms.txt](https://docs.dapta.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.dapta.ai/dapta-mcp/privacy.md). # Privacy & Data Handling What the Dapta connector for supported AI clients collects, how Dapta uses it, and how long it is kept. This page covers the Dapta MCP, the connector at `https://mcp.dapta.ai/mcp`. It adds to the [Dapta Privacy Policy](https://dapta.ai/privacy-policy), which applies to your whole Dapta account and to the data in your workspace. *** ## What the connector collects The connector does not receive your full conversation with the connected AI client. It receives only the tool calls the connected AI client makes for you: the name of the tool and its arguments. For every tool call, Dapta records: | Data | Detail | | --------------------------------------- | -------------------------------------------------------------------------------------------------- | | **Who and where** | Your Dapta user id and email, the workspace id, and the id of the API key when one is used | | **What ran** | Tool name, time, duration, and whether it succeeded or failed, with the error message if it failed | | **Arguments of tools that read** | The arguments as sent: ids, filters, dates, page sizes | | **Arguments of tools that change data** | Only a fingerprint (a hash) of the arguments, not their content | When your AI client prepares a change, Dapta also stores the **pending change**: what will be written and the preview you were shown. This is what makes the confirm step apply exactly what you saw. Dapta also collects **usage statistics**: tool name, duration, error category, the name and version of the app you connected from, and internal ids. They carry no names, emails or workspace content. The connector does not record what the tools return to the connected AI client. *** ## How Dapta uses it * To run the connector: checking who you are, which workspace you are in, and applying the changes you confirm. * For security: an audit trail of what was done through the connector, and usage limits. * To fix problems and improve the connector. *** ## How long it is kept | Data | Kept for | | ------------------------------------------ | ---------------------------------------------------------------------------- | | **Pending changes** | They expire 15 minutes after the preview and are deleted about one day later | | **Tool call records and usage statistics** | No fixed expiry. They are kept until you ask Dapta to delete them | To have this data deleted, write to . *** ## What your AI provider receives Tool results are sent to the AI client you connected. When you use Dapta through ChatGPT or Codex, OpenAI receives those results and handles them under your agreement with OpenAI. When you use Claude, Anthropic receives them and handles them under your agreement with Anthropic. When you use another client, such as Cursor, its provider receives them under your agreement with that provider. Results contain the workspace data you requested, such as agent configurations, call records and transcripts, text conversations, automation flows and execution history, or database rows. Depending on the selected tool and your workspace content, results may include names, contact details and other personal data stored in your workspace. *** ## Contact Privacy, support and security reports: --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.dapta.ai/dapta-mcp/privacy.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.