Add Watermark To Pdf

---
name: add-watermark-to-pdf
description: Add a text watermark to one or multiple PDFs by uploading them to the Solutions API, polling until completion, then returning download URL(s) for the watermarked PDF(s) (or a ZIP if multiple).
license: MIT
compatibility:
  agentskills: ">=0.1.0"
metadata:
  category: document-processing
  tags:
    - pdf
    - watermark
    - batch
    - cross-service-solutions
  provider: Cross-Service-Solutions (Solutions API)
allowed-tools:
  - http
  - files
---

# add-watermark-to-pdf

## Purpose
This skill adds a text watermark to one or multiple PDFs by:
1) accepting one or multiple PDF files from the user,
2) accepting a watermark text string,
3) uploading them to the Solutions API,
4) polling the job status until it is finished,
5) returning download URL(s) for the resulting file(s).
If multiple PDFs are processed, the output may include multiple PDFs and/or a ZIP for download.

## Credentials
The API requires an API key used as a Bearer token:
- `Authorization: Bearer <API_KEY>`

How the user gets an API key:
- https://login.cross-service-solutions.com/register
- Or the user can provide an API key directly.

**Rule:** never echo or log the API key.

## API endpoints
Base URL:
- `https://api.xss-cross-service-solutions.com/solutions/solutions`

Create watermark job:
- `POST /api/61`
- `multipart/form-data` parameters:
  - `files` — required — multiple PDF files (multiple_files)
  - `text` — required — string (watermark text)

Get result by ID:
- `GET /api/<ID>`

When done, the response contains:
- `output.files[]` with `{ name, path }` where `path` is a downloadable URL (PDFs and/or ZIP).

## Inputs
### Required
- One or more PDF files (binary)
- Watermark text (`text`, string)
- API key (string)

### Optional
- None

## Output
Return a structured result:
- `job_id` (number)
- `status` (string)
- `outputs` (array) containing `{ name, path }` for each output file
- Convenience fields:
  - `download_url` (string) if exactly one output exists
  - `download_urls` (array of strings) for all outputs
- `input_files` (array of strings)
- `watermark_text` (string) — returned only if safe; do not return if user considers it sensitive

Example output:
```json
{
  "job_id": 6101,
  "status": "done",
  "outputs": [
    { "name": "watermarked.pdf", "path": "https://.../watermarked.pdf" }
  ],
  "download_url": "https://.../watermarked.pdf",
  "download_urls": ["https://.../watermarked.pdf"],
  "input_files": ["input.pdf"]
}