PandaDoc API Patterns

PandaDoc MCP Tools & API Patterns

Overview

PandaDoc provides a hosted MCP server at https://developers.pandadoc.com/mcp for AI tool integration. The MCP server provides direct API access to PandaDoc's document automation platform, documentation search, and code generation assistance. This skill covers MCP server connection, authentication, the complete tool reference, rate limiting, error handling, and best practices.

Connection & Authentication

MCP Server

PandaDoc hosts an official MCP server. Authentication uses an API key passed via the Authorization header:

1. Log into app.pandadoc.com
2. Navigate to Settings > Integrations > API
3. Generate an API key

MCP Server URL: https://developers.pandadoc.com/mcp

Required Header:

| Header | Value | Description | |--------|-------|-------------| | Authorization | API-Key <key> | API key from PandaDoc Settings |

Authentication Modes

| Mode | Auth Required | Use Case | |------|---------------|----------| | Documentation search | No | Browsing PandaDoc API docs and guides | | Code generation assistance | No | Generating code snippets for PandaDoc integration | | Live API calls | Yes | Creating documents, sending for signature, checking status |

Note: Documentation search and code generation work without an API key. All live API operations (creating documents, managing templates, sending for signature) require a valid API key.

Environment Variables

export PANDADOC_API_KEY="your-api-key"

Claude Desktop Configuration

{
  "mcpServers": {
    "pandadoc": {
      "command": "npx",
      "args": [
        "-y", "mcp-remote",
        "https://developers.pandadoc.com/mcp",
        "--header", "Authorization:API-Key YOUR_API_KEY"
      ]
    }
  }
}

Complete MCP Tool Reference

Document Tools

| Tool | Description | Key Parameters | |------|-------------|----------------| | pandadoc-list-documents | List and filter documents | status, q (search query), tag, count, page, order_by, template_id, folder_uuid | | pandadoc-get-document | Get a single document's details | id (required) | | pandadoc-get-document-status | Get document status | id (required) | | pandadoc-create-document | Create a document from template | template_uuid, name, recipients, tokens, fields, pricing_tables, folder_uuid | | pandadoc-send-document | Send a document for signature | id (required), message, subject, silent | | pandadoc-download-document | Download a completed document | id (required) |

Template Tools

| Tool | Description | Key Parameters | |------|-------------|----------------| | pandadoc-list-templates | List available templates | q (search query), count, page, tag, folder_uuid | | pandadoc-get-template | Get template details | id (required) |

Recipient Tools

| Tool | Description | Key Parameters | |------|-------------|----------------| | pandadoc-add-recipient | Add a recipient to a document | document_id (required), email, first_name, last_name, role, signing_order |

Documentation Tools

| Tool | Description | Key Parameters | |------|-------------|----------------| | pandadoc-search-docs | Search PandaDoc API documentation | query (required) | | pandadoc-get-code-sample | Generate code samples | endpoint, language |

Pagination

Page-Based Pagination

List tools use 1-based page pagination with configurable count:

Pagination Parameters:

| Parameter | Description | Default | Max | |-----------|-------------|---------|-----| | page | Page number (1-based) | 1 | - | | count | Results per page | 50 | 100 |

Iterating Through All Pages

To fetch all results, call the list tool repeatedly, incrementing page from 1 until the number of results returned is less than count:

1. Call the tool with page=1 and count=100
2. If the response contains 100 results, call again with page=2
3. Continue until fewer than 100 results are returned
4. Collect results from each response

Filtering & Sorting

Document Filters

| Parameter | Values | Description | |-----------|--------|-------------| | status | document.draft, document.sent, document.completed, document.viewed, document.waiting_approval, document.approved, document.rejected, document.waiting_pay, document.paid, document.voided, document.declined, document.expired | Filter by document status | | q | string | Search documents by name | | tag | string | Filter by document tag | | template_id | UUID | Filter by source template | | folder_uuid | UUID | Filter by folder | | order_by | date_created, date_modified, name, date_status_changed | Sort field |

Template Filters

| Parameter | Values | Description | |-----------|--------|-------------| | q | string | Search templates by name | | tag | string | Filter by template tag | | folder_uuid | UUID | Filter by folder |

Response Format

Single Document:

{
  "id": "msFYActMfJHqNTKH9tcPFa",
  "name": "Acme Corp - Managed Services Agreement",
  "status": "document.sent",
  "date_created": "2026-01-15T10:30:00.000000Z",
  "date_modified": "2026-01-16T14:22:00.000000Z",
  "expiration_date": "2026-02-15T00:00:00.000000Z",
  "version": "2",
  "recipients": [
    {
      "email": "john@acme.com",
      "first_name": "John",
      "last_name": "Smith",
      "role": "Client",
      "signing_order": 1,
      "has_completed": false
    }
  ],
  "tokens": [
    {
      "name": "Client.Company",
      "value": "Acme Corporation"
    }
  ],
  "grand_total": {
    "amount": "2500.00",
    "currency": "USD"
  }
}

Paginated Document List:

{
  "results": [
    {
      "id": "msFYActMfJHqNTKH9tcPFa",
      "name": "Acme Corp - Managed Services Agreement",
      "status": "document.sent",
      "date_created": "2026-01-15T10:30:00.000000Z"
    }
  ]
}

Rate Limiting

Rate Limit Details

| Plan | Limit | |------|-------| | Business | 300 requests per minute | | Enterprise | 600 requests per minute |

When rate limited, the API returns HTTP 429. Wait before retrying. Use exponential backoff for automated retries.

Rate Limit Best Practices

1. Batch operations - Minimize total API calls by using filters and pagination efficiently
2. Cache template data - Templates change infrequently; cache for minutes to hours
3. Avoid polling - Use webhooks where possible instead of repeatedly checking document status
4. Stagger requests - When processing multiple documents, add small delays between calls

Error Handling

Common Errors

| Error | HTTP Code | Cause | Resolution | |-------|-----------|-------|------------| | Unauthorized | 401 | Invalid or missing API key | Check PANDADOC_API_KEY and regenerate if needed | | Forbidden | 403 | Insufficient permissions | Verify API key has required access scope | | Not Found | 404 | Document/template ID does not exist | Verify the resource ID | | Rate Limited | 429 | Too many requests | Wait and retry with exponential backoff | | Validation Error | 422 | Invalid request parameters | Check required fields and parameter formats | | Server Error | 500 | PandaDoc internal error | Retry after a brief delay |

Troubleshooting MCP Connection

1. Verify API key - Ensure the API key is valid and not revoked
2. Check URL - MCP server URL must be https://developers.pandadoc.com/mcp
3. Test with docs search - Try pandadoc-search-docs (no auth required) to verify connectivity
4. Test with a simple call - Try pandadoc-list-documents with count=1 to verify authentication
5. Regenerate key - If authentication fails, generate a new key from PandaDoc Settings > API

Best Practices

1. Use template-based creation - Always create documents from templates for consistency
2. Filter server-side - Use status, q, and `tag