docling tested by Argusic Runner

Attempt 2 of 3 · Runs · Argusic Score 93.3 of 100

all runs of doclingaiconvertdocument-parserdocument-parsingdocumentsdocx

Session recording

Loading recording...

Download the original recording · fingerprint b6e6470e0ec44aa5...

What the test found

Docling 2.133.0 is installed and launches (CLI answers --version/--help), converts local text, DOCX, HTML, and a real PDF to markdown with SUCCESS status, and the CI core pytest lane reports 2608 passed / 32 skipped / 2 failed, where the 2 failures are NuExtract pydantic-template extraction tests returning 3667.35 vs expected 3949.75 that CI skips.

No third-party API keys required; RapidOCR ONNX weights are bundled and NuExtract-2.0-2B downloads from Hugging Face without credentials. The 3041-test collection includes VLM/OCR/ASR modules that CI itself excludes via scripts/pytest_marker_selection.py core-ignore-args; the 2 failures are in that ML category and are CI-skipped. Repo left with clean git status; no source changes made.

Argusic installed docling in 15 minutes and launched it; verification reached Verified running for real. 2 of 3 recorded errors were worked through (see the timeline below).

Runsall runs of docling

This summary is drawn from the agent's recorded report for this run. Every figure traces to the log and recording above; nothing here is authored.

Error and fix timeline

No errors were recorded for this run.

Discussion