We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/caffeinebounce/docuseal-mcp-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
README.md•2 KiB
# docuseal-mcp-server
MCP server for DocuSeal (self-hosted or cloud) using stdio transport.
Package name: `@caffeinebounce/docuseal-mcp-server`
## Features
Implements these MCP tools:
### Templates
- `list_templates` — list templates (paginated)
- `get_template` — get template by ID
- `create_template_from_pdf` — upload PDF from file path or base64
### Submissions
- `create_submission` — create signature request from template
- `list_submissions` — list submissions (paginated + status filter)
- `get_submission` — get submission details
- `get_submission_documents` — get documents for a submission (optional local download)
### Submitters
- `list_submitters` — list submitters (optionally by submission)
- `get_submitter` — get submitter details
- `update_submitter` — update submitter fields (including reminder-style flags like `send_email`)
## Requirements
- Node.js 18+
- A DocuSeal instance with API enabled
- API key with required permissions
## Environment Variables
- `DOCUSEAL_URL` — base URL for DocuSeal (default: `http://localhost:3030`)
- `DOCUSEAL_API_KEY` — API key for `X-Auth-Token` header (**required**)
Example:
```bash
export DOCUSEAL_URL="http://localhost:3030"
export DOCUSEAL_API_KEY="<your-api-key>"
```
## Install and Build
```bash
npm install
npm run build
```
## Run
```bash
npm start
```
Or in development:
```bash
npm run dev
```
## MCP Client Configuration (example)
```json
{
"mcpServers": {
"docuseal": {
"command": "node",
"args": ["/absolute/path/to/docuseal-mcp-server/dist/index.js"],
"env": {
"DOCUSEAL_URL": "http://localhost:3030",
"DOCUSEAL_API_KEY": "<your-api-key>"
}
}
}
}
```
## Notes
- API calls use `fetch` and include `X-Auth-Token`.
- API errors include HTTP status + response body in tool output for debugging.
- `get_submission_documents` can optionally download returned document URLs to disk using:
- `download: true`
- optional `output_dir` (default: `./downloads/submission_<id>`)