How It Works

The DocEndorse MCP server enables users to initiate and manage document execution and electronic signature (e-signature) workflows through compatible AI assistants, such as ChatGPT and Claude.

To get started, users must connect to the DocEndorse MCP server at https://mcp.docendorse.com/mcp and authenticate with the DocEndorse platform. A DocEndorse account is required. If you do not already have an account, follow the account setup instructions here.

Once connected and authenticated, users can initiate supported document workflows through natural language instructions. The process follows these steps.

2.1 Step 1 – Provide an Instruction

The user describes the document or e-signature task they want to complete in their MCP-compatible AI assistant.

For example, the user might ask the assistant to prepare a document for signature, find an existing template, send a signature request, or check the status of a document.

Providing a Document

If the workflow requires a document, the user can provide one in either of the following ways:

Option A: Use an existing DocEndorse document

If the document has already been uploaded to DocEndorse, the user can ask the assistant to search for it and use it in the requested workflow.

Option B: Upload a document through the MCP client

If the document is not already available on DocEndorse, the user may be able to upload it directly through the MCP client and instruct the assistant to pass it to the DocEndorse AI Agent.

This approach depends on the MCP client's capabilities and configuration. The client must support the necessary file-handling operations and, where required, script execution and network access. The docendorse.com domain must also be allowlisted where the client's security configuration requires it.

Option C: Upload a document through the DocEndorse AI Agent

Alternatively, the user can upload the document directly through the DocEndorse AI Agent's composer after the workflow has been handed off to the agent.

This option allows users to provide the document within the AI Agent's interface rather than uploading it through the MCP client.

Using a Template

If the workflow requires a template, the user must locate the appropriate template before it can be used.

The template can be searched for through the MCP client or located after the workflow has been handed off to the DocEndorse AI Agent.

Template use and document preparation are always handled by the DocEndorse AI Agent.

2.2 Step 2 – The Assistant Invokes DocEndorse

Once the user has provided an instruction and any required information, the AI assistant uses the appropriate MCP tool to interact with DocEndorse.

Depending on the request, the assistant may retrieve document or template information, check signature status, initiate a supported operation, or hand off the workflow to the DocEndorse AI Agent.

The MCP server acts as the communication layer between the AI assistant and DocEndorse, allowing the assistant to access the supported capabilities exposed by the server.

2.3 Step 3 – The DocEndorse AI Agent Handles Advanced Workflows

All template creation and document preparation workflows are handled by the DocEndorse AI Agent.

When a user requests one of these operations, the MCP server hands the workflow off to the AI Agent, transferring the relevant context automatically.

The AI Agent then performs the necessary document processing. Depending on the task, this may include:

• Reviewing the supplied document.
• Detecting editable fields and identifying information that needs to be completed.
• Identifying signatory roles and the parties involved in the signing process.
• Identifying and configuring signature fields.
• Presenting detected fields and signatory roles for the user to review.
• Allowing the user to modify, add, or replace fields and signatory roles.

The user can review the proposed document structure and make any necessary adjustments before proceeding.

This handoff allows users to initiate workflows through their preferred AI assistant while using the DocEndorse AI Agent for the document processing and interactive preparation tasks it is designed to handle.

2.4 Step 4 – DocEndorse Executes the Requested Operation

DocEndorse processes the request and performs the relevant supported operation.

Depending on the workflow, this may involve retrieving document information, creating or preparing a document through the AI Agent, initiating an e-signature request, or performing another supported action.

The result is returned through the appropriate interface, allowing the user to review the outcome and continue with any necessary next steps.

For document execution workflows, the process may continue after the initial request while signatories review and sign the document.

2.5 Step 5 – Track and Manage Signature Requests

After an e-signature request has been initiated, users can use the supported DocEndorse capabilities to check document status and manage outstanding signature requests.

Depending on the available capabilities, users can:

• Retrieve the current status of a document or signature request.
• Identify signature requests that still require action.
• Send reminders or follow-ups to signatories.
• Resend existing signature requests when necessary.

These capabilities allow users to monitor document execution and manage follow-up actions without having to restart the original workflow.

The information returned reflects the document and signature status available in DocEndorse when the request is processed.

2.6 Summary

The DocEndorse MCP server provides a conversational way to access supported document execution and e-signature capabilities through compatible AI assistants.

The process begins with a user instruction, followed by an MCP request to DocEndorse. For template creation and document preparation, the workflow is always handed off to the DocEndorse AI Agent, which performs the necessary document processing and guides the user through preparation.

Once the relevant operation has been completed, users can continue managing the document lifecycle through supported capabilities, including signature tracking, reminders, and resending signature requests.