Self-hosted server

Turn features on or off

Remove tools you don't want people to use, or that your server can't run. A removed tool is greyed out in the app (or hidden, see below) and its API endpoint is switched off for everyone on the server.

Remove tools#

In Settings → Server → System → Endpoint Management, pick tools under Disabled Endpoints or whole groups under Disabled Endpoint Groups, then select Save Changes and Restart Now.

Or set them in configuration. As environment variables, separate values with commas:

yaml
endpoints:
  toRemove:
    - sign
    - add-watermark
  groupsToRemove:
    - LibreOffice
bash
ENDPOINTS_TOREMOVE=sign,add-watermark
ENDPOINTS_GROUPSTOREMOVE=LibreOffice

IDs and group names are case-sensitive. The same IDs control which tools the MCP server exposes, and a tool removed here is never available over MCP.

Hide or grey out removed tools#

By default, removed tools stay visible but greyed out. To hide them instead, turn on Hide unavailable tools by default and Hide unavailable conversions by default under Settings → Server → System → User Preference Defaults (ui.defaultHideUnavailableTools and ui.defaultHideUnavailableConversions). People can change this in their own settings.

Tool groups#

Removing one of these groups removes every tool in it.

Group Tool IDs
PageOps add-page-numbers, auto-rotate-pdf, auto-split-pdf, booklet-imposition, crop, extract-pages, merge-pdfs, multi-page-layout, overlay-pdf, pdf-to-single-page, rearrange-pages, remove-pages, rotate-pdf, scale-pages, split-by-size-or-count, split-pages, split-pdf-by-chapters, split-pdf-by-sections
Convert cbz-to-pdf, eml-to-pdf, file-to-pdf, html-to-pdf, img-to-pdf, json-to-pdf, markdown-to-pdf, pdf-to-cbz, pdf-to-csv, pdf-to-epub, pdf-to-html, pdf-to-img, pdf-to-json, pdf-to-markdown, pdf-to-pdfa, pdf-to-presentation, pdf-to-rtf, pdf-to-text, pdf-to-ua, pdf-to-vector, pdf-to-video, pdf-to-word, pdf-to-xml, url-to-pdf, vector-to-pdf
Security accessibility-report, add-password, add-stamp, add-watermark, auto-redact, cert-sign, change-permissions, redact, remove-cert-sign, remove-password, sanitize-pdf, sign, timestamp-pdf, unlock-pdf-forms, validate-compliance, validate-signature, verify-pdf
Other add-attachments, add-image, batch-process-attachments, compare, delete-fields, edit-table-of-contents, extract-images, extract-single-attachment, fields, fill, flatten, get-info-on-pdf, list-attachments, modify-fields, multi-tool, ocr-pdf, remove-annotations, remove-blanks, replace-invert-pdf, text-editor-pdf, update-metadata, view-pdf
Advance adjust-contrast, auto-rename, compress-pdf, extract-image-scans, overlay-pdf, repair, scanner-effect
Automation automate, handleData, pipeline
DocParse ingest
DeveloperTools show-javascript
DeveloperDocs dev-api-docs, dev-airgapped-docs, dev-folder-scanning-docs, dev-sso-guide-docs (links to developer guides in the app)

Tools outside these groups, such as pdf-to-cbr or remove-image-pdf, can still be removed by ID. A tool's ID is the last part of its API path, and conversions use <from>-to-<to>, so /api/v1/convert/pdf/img is pdf-to-img. The API docs at /swagger-ui/index.html on your server list every path.

Dependency groups#

These groups are named after what a tool needs to run. When a program is missing, Stirling PDF removes its group at startup. Tools with another way to run, such as compress-pdf, stay available.

Group Tools that use it
LibreOffice file-to-pdf, pdf-to-html, pdf-to-pdfa, pdf-to-presentation, pdf-to-rtf, pdf-to-word, pdf-to-xml
Unoconvert file-to-pdf
Python extract-image-scans, file-to-pdf, html-to-pdf, url-to-pdf
OpenCV extract-image-scans
Weasyprint eml-to-pdf, html-to-pdf, markdown-to-pdf, url-to-pdf
Calibre pdf-to-epub
qpdf compress-pdf, repair
Ghostscript compress-pdf, crop, pdf-to-vector, repair, replace-invert-pdf, scanner-effect, vector-to-pdf
ImageMagick compress-pdf
OCRmyPDF ocr-pdf (runs with either OCRmyPDF or tesseract)
tesseract auto-rotate-pdf, ocr-pdf
Pdftohtml pdf-to-html, pdf-to-markdown
veraPDF accessibility-report, pdf-to-ua, validate-compliance, verify-pdf
Java Built-in tools that need no extra program, such as merge-pdfs, split-pages, compress-pdf and pdf-to-img. Removing this group turns off most of the app.
Javascript adjust-contrast, compare, rearrange-pages, sign, text-editor-pdf
rar pdf-to-cbr (see PDF to CBR conversion)
CLI compress-pdf, extract-image-scans, file-to-pdf, html-to-pdf, ocr-pdf, pdf-to-html, pdf-to-pdfa, pdf-to-presentation, pdf-to-rtf, pdf-to-word, pdf-to-xml, repair, url-to-pdf. The ultra-lite image removes this group.

Tools that are off by default#

url-to-pdf stays off unless you turn on Enable URL to PDF under Settings → Server → Advanced → Feature Flags (system.enableUrlToPDF). It fetches web pages from your server, so only enable it for trusted internal use.

Auto Form Detection (form-detection) only runs on the standard and fat Docker images, which ship with a detection model. To install it, open Settings → Server → System → AI Form Detection and select Install next to the model. For a server without internet access, follow the steps under Offline / air-gapped install on the same card. To turn the tool off, switch off Enable feature there (formDetection.enabled).