Run a Docling API server for agent document conversion
Deploy Docling API so agents can convert PDFs, Office files, images, HTML, CSV, and other documents into Markdown through sync, batch, and async HTTP endpoints.
npx skills add agentskillexchange/skills --skill run-a-docling-api-server-for-agent-document-conversion
https://github.com/drmingler/docling-api, run poetry install, configure Redis in .env, start Redis, then run poetry run uvicorn main:app --reload --port 8080 and call endpoints such as /documents/convert.Use Docling API when an agent needs a repeatable document-ingestion service instead of one-off local parsing. The operator deploys the FastAPI, Celery, and Redis service, then agents call the conversion endpoints to turn documents into Markdown, extract tables and images, run OCR, and track asynchronous batch jobs. This is scoped to running a document conversion API that agents can call in ingestion and RAG-preparation workflows; it is not a generic Docling library listing or a document management product.
What this skill actually does
Inputs and prerequisites: Python 3.12, Poetry, Redis, FastAPI, Celery, optional GPU runtime.
Setup notes: Clone https://github.com/drmingler/docling-api , run poetry install , configure Redis in .env , start Redis, then run poetry run uvicorn main:app –reload –port 8080 and call endpoints such as /documents/convert .
Source and verification boundary: use https://github.com/drmingler/docling-api as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.