LM-Kit OneDocs2026.8.10lm-kit.com
Documents

The PDF Toolbox

PDF is where documents go to become permanent, so the server treats it as a first-class surface: a family of /lmkit/v1/pdf-* endpoints deep enough that each concern has its own chapter below. All of them take the same input contract (base64 or an uploaded file id), return file ids that chain into the next call, accept a password where encrypted input is plausible, and ride the jobs contract when work runs long. This page is the map.


1The chapters#

Chapter The concern Endpoints
Redaction Content destruction with proof: selectors, scope toggles, forensic counters, and searchable verification pdf-redact
Searchable PDFs and OCR Scans that become text-bearing documents: the invisible overlay, page policies, compression, direct PDF/A output pdf-ocr, document-ocr
PDF/A Conversion and Validation The archival workflow: repair-first conversion and the independent conformance auditor pdf-to-pdfa, document-validation
Assembly and Page Surgery Deterministic mechanics: merge, split, reorder, rotate, unlock, and the image bridge both ways pdf-merge, pdf-split, pdf-edit, pdf-unlock, image-to-pdf, pdf-to-images
Search, Layout, and Inspection Reading without transforming: three-mode search with coordinates, page geometry, and triage facts pdf-search, pdf-layout, pdf-info
Digital Signatures The cryptographic life story: PAdES signing, four-axis verification, document timestamps, LTV archival pdf-sign, pdf-verify-signatures, pdf-timestamp, pdf-ltv

2One workshop, composable#

The chapters are stages of each other's workflows, which is the point of shipping them as one API. The recurring compositions:

3Stated plainly#

  • Six chapters, one contract: shared inputs, chainable file ids, honest reports, and password handling where reality requires it.
  • The toolbox operates on one PDF at a time; questions across thousands of documents belong to Search.
  • If a workflow touches PDFs at all, its verbs are here; start from the chapter that owns the concern.