BioResearcherPubMed weekly — updatefiles sample parsed into an Excel workbook
PASS · rubric SATISFIED56sprompt: enartifact: en
Prompt (verbatim)
In {DATA_DIR}/pubmed-sample.xml.gz you have a trimmed PubMed updatefiles sample. Load the bioresearcher-pubmed-weekly skill and parse this file into combined.xlsx in the working directory, then report the sheet names, row counts, and the deleted PMIDs.
What happened
- Parsed a trimmed weekly updatefiles archive (6 <PubmedArticle> entries + a <DeleteCitation> block) via the skill's pure-Python streaming parser.
- Output verified against the committed expected-summary.json: 6 article rows, deleted PMIDs 99999991/99999992 present.
- Rubric flag verify-combined-xlsx adjudicated SATISFIED post-run (recorded in result.json).
Evidence
| Sheet | Rows | Content |
| PubMed Articles | 6 (+ header) | PMIDs 12345678, 87654321, 31000001, … (incl. DOI/journal/author columns) |
| Deleted PMIDs | 2 (+ header) | 99999991, 99999992 |
Objective checks
checks passed: 1/1 · one rubric flag adjudicated SATISFIED post-run (recorded in result.json)
| Fan-out mode | n/a — single procedural parse, no research fan-out |
|---|---|
| Provenance | 3380cc6 · opencode 1.18.29 |
Replay
node demos/run-demo.mjs --only agent-pubmed-weekly-en --publishManual-run only — this replay spends LLM tokens (real opencode CLI + network).
In the repository: README · transcript · result.json · artifact directory (permalink)