workshop-paper-reviewlisted

# Workshop Paper Review Use this skill to produce a polished OpenReview workshop review (target: ~400 words English, 5 sections, single rating). The skill embeds hard-won conventions from a 4-paper review batch and is opinionated about the anti-patterns that creep into LLM-drafted reviews. It is **not** a friendly read-through; it is structured for honest scoring, fact-grounded weaknesses, rebuttal-friendly questions, and OpenReview submission style. ## When to use - Reviewing an ICML / NeurIPS / ICLR workshop paper (4–8 page submissions, scope-clear venues). - Drafting a Korean working review for personal calibration before producing the English OpenReview submission. - Fact-checking an existing review draft against the source PDF. - Polishing a draft to remove AI-writing patterns, severity tags, vague compliments, and reviewer-centric phrasing. If the user wants an adversarial pre-submission audit on their *own* draft (not someone else's), use `adversarial-review` instead. ## Inputs to confirm Ask only for what is missing: - path to the source PDF - workshop venue (only matters for CFP scope sanity; otherwise generic OpenReview workshop form is assumed) - output language(s): Korean only, English only, or both (default: both — Korean working draft + English final) - output directory (default: same directory as the PDF) - whether a v3-style full internal review is also wanted alongside the short OpenReview submission (default: no — submission only) ## OpenReview works