Convert PDF and images to 10 structured formats locally via a Python CLI wrapper around the ComPDFKit SDK.
Design & media
PDF Extract
Run 50+ PDF and document processing operations through the ComPDF Cloud API from one agent skill.
What it does
The skill wraps calls to ComPDF Cloud's REST API for PDF format conversion (PDF↔Word, Excel, Image, Markdown), page editing (merge, split, watermark, compress), OCR, and AI document extraction or parsing. Each operation uploads the user's file to ComPDF servers, so explicit consent is required before transmission, and results come back synchronously with a download link that expires at 24:00 the next day. The skill handles API key storage on user opt-in, supports both international and mainland China endpoints, and enforces the service's free-tier limit of 200 files per 30 days.
When to use it
- Convert a PDF to Word, Excel, Image, or Markdown
- Merge, split, watermark, or compress PDFs
- Run OCR or AI extraction on scanned documents
- Process an encrypted PDF by supplying its password
The skill document
PDF Extract
Process PDF files through ComPDF Cloud REST API. Supports 50+ document processing operations.
Official documentation:
When to Run
- User requests to convert file format (e.g., "convert this PDF to Word", "convert Excel to PDF")
- User requests to edit PDF pages (e.g., "merge these two PDFs", "delete page 3", "rotate PDF")
- User requests to add or remove watermarks from PDF
- User requests to compress PDF files
- User requests OCR recognition of scanned documents or text in images
- User requests AI extraction or parsing of document content
- User requests to extract tables from images
- User requests batch processing of multiple document files
- User requests to compare differences between two PDF documents
- User mentions ComPDF, compdf, or related keywords
Workflow
Step 1 — Obtain API Key
Check whether config/public_key.txt exists and contains a non-empty value.
- If the file exists and is non-empty: use the stored key (trim whitespace).
- If the file is missing or empty: ask the user for their ComPDF API Public Key. Inform them it can be obtained at . After the user provides the key, ask whether they would like to save it locally for future sessions.
- If the user agrees, write the key to
config/public_key.txt. - If the user declines, use the key for the current session only without saving.
- If the user agrees, write the key to
The key file is not included in the published skill package. It is created at runtime only when the user explicitly opts in. The user may delete
config/public_key.txtat any time to revoke local storage.
Step 2 — Confirm External Upload Intent
Before uploading any file, explicitly inform the user:
⚠️ External Upload Confirmation Required
Your file will be uploaded to ComPDF's servers (api-server.compdf.com or api-server.compdf.cn) for processing. Please confirm that:
- You consent to uploading this file to external servers.
- The file does not contain highly sensitive or confidential data, or you accept the associated risk.
- You have reviewed ComPDF's Privacy Policy at .
Only proceed with the upload after receiving explicit user confirmation.
Step 3 — Determine Base URL
Ask or infer the user's network environment:
| Environment | Base URL |
|---|---|
| International | https://api-server.compdf.com/server/v2 |
| Mainland China | https://api-server.compdf.cn/server/v2 |
Step 4 — Select Tool
Look up the executeTypeUrl for the user's task in references/tool-list.md.
Quick reference for common operations:
| Operation | executeTypeUrl |
|---|---|
| PDF → Word | pdf/docx |
| PDF → Excel | pdf/xlsx |
| PDF → Image | pdf/img |
| PDF → Markdown | pdf/markdown |
| Word → PDF | docx/pdf |
| Merge PDF | pdf/merge |
| Split PDF | pdf/split |
| Add Watermark | pdf/addWatermark |
| PDF Compression | pdf/compress |
| OCR | documentAI/ocr |
| AI Document Extraction | idp/documentExtract |
| AI Document Parsing | idp/documentParsing |
For the full list of 50+ tools, see references/tool-list.md.
Step 5 — Build Parameters (optional)
If the selected tool supports custom parameters, look up its JSON schema in references/parameters.md. Parameters are passed as a JSON string in the parameter form-data field. If omitted, server defaults apply.
Step 6 — Send Request
After the user has explicitly confirmed the external upload:
- Send a
POSTrequest to{baseUrl}/process/{executeTypeUrl}. - Include the user-provided API key in the
x-api-keyheader for the current session only. - Send the selected file as multipart form-data.
- Include the
parameterfield only when the selected tool supports custom parameters. - Use the synchronous
/process/endpoint so the complete result is returned in a single response.
Do not send the request until the user has confirmed that the file may be transmitted to ComPDF Cloud.
Step 7 — Handle Response
1. Check code field — "200" means success; anything else is an error.
2. Check taskStatus (should be TaskFinish for synchronous calls):
| Status | Meaning | Action |
|---|---|---|
TaskFinish | Processing complete | Proceed to download |
TaskProcessing | Still processing internally | Notify user; suggest retrying shortly |
TaskOverdue | Timed out | Retry or split into smaller tasks |
3. Extract download link from fileInfoDTOList[].downloadUrl.
4. Warn the user: download links expire at 24:00 the next day.
5. On failure: read failureCode and failureReason, then look up troubleshooting advice in references/error-codes.md.
6. On quota exhaustion (code = "06001"): inform the user:
For more credits, please visit
Auxiliary Endpoints
| Purpose | Method & Path |
|---|---|
| List supported tools | GET {baseUrl}/tool/support |
| Check remaining credits | GET {baseUrl}/asset/info |
| List tasks | GET {baseUrl}/task/list?page=1&size=10 |
| Close a task | POST {baseUrl}/task/closeTask?taskId={taskId} |
Output Format
On success:
Processing complete!
File: {fileName} → {downFileName}
Status: {taskStatus}
Time taken: {convertTime}ms
Original size: {fileSize} bytes
Result size: {convertSize} bytes
Download link: {downloadUrl}
⚠️ The download link will expire at 24:00 tomorrow, please save it promptly.
On failure:
Processing failed.
Error code: {failureCode}
Reason: {failureReason}
Suggestion: {troubleshooting suggestion from references/error-codes.md}
Critical Rules
- HTTP 200 ≠ success — always check
codeandtaskStatusin the JSON body. - Max 5 files per task — split into multiple tasks if more files are needed.
- Free tier limit — 200 files within 30 days.
- File type auto-detection — the API detects uploaded file types; no need to rename extensions.
- Encrypted PDFs — provide the password in the
passwordform-data field (separate fromparameter). - China domain — mainland China users must replace
compdf.comwithcompdf.cn. - User-controlled API Key storage — the key file (
config/public_key.txt) is never shipped with the skill package. It is created at runtime only when the user explicitly opts in. The user may delete it at any time. - External upload confirmation — always obtain explicit user consent before uploading files to ComPDF servers.
License & Copyright
Copyright © 2014-2026 PDF Technologies, Inc., a KDAN Company. All Rights Reserved.
ComPDF and ComPDFKit are trademarks of PDF Technologies, Inc., a KDAN Company.
This skill package is licensed under the Apache License 2.0. See LICENSE.txt for the full license text.
The ComPDF Cloud API is a commercial service provided by PDF Technologies, Inc. Use of the API is subject to the ComPDF Terms of Service and Privacy Policy.
Related skills
Convert PDF or image files to 10 editable formats (Word, Excel, PPT, HTML, and more) via ComPDFKitConversion.
Files and audits incoming invoices — OCR, duplicate checks, VAT splits, and a searchable local archive.
Generate polished PDF and DOCX documents from prompts — resumes, contracts, reports, proposals, and more.
Edit, merge, split, rotate, and convert PDFs from the command line with the ComPDFKit CLI engine.
Process PDFs through the ComPDF Cloud REST API with 50+ operations: convert, merge, split, watermark, OCR, and more.