Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/cikm-writing-style
Use when revising a CIKM manuscript for the tri-community readership — writing an opening that lands with IR, data-mining, and knowledge-management reviewers simultaneously, compressing into appendix-inclusive page budgets, keeping claims inside the evidence, and maintaining double-blind and disclosure-compliant prose.
Passed 23 of 26 test cases with the skill, and 12 of 26 without it.
and 13 more in the full report
11 fixed12 passed either way3 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 11 test cases the skill fixed. Same model, same prompt, with and without the skill.
During a first-page scan audit of a manuscript, we find that the paper describes a single-lane IR model without any interaction across mining or database boundaries. What workflow recommendation should be given before doing further text polishing?
| Checked | Without skill | With skill |
|---|---|---|
| Recommends re-running topic selection before further text polishing. | Missed | Passed |
The instructions your model reads when the skill loads.
A CIKM paper is read in three accents at once. The IR reviewer scans for the task and the evaluation; the mining reviewer scans for the mechanism and its delta; the KM/database reviewer scans for the data reality — schemas, scale, heterogeneity, noise. Style at this venue means making all three scans succeed in the first page, inside a budget where appendices offer no escape (10 pages including them for 2026 full papers).
Order the opening so each lane hits its anchor early:
If the paper cannot state point 2, it is a single-lane paper and
cikm-topic-selection should re-run before more polishing.
Cross-community words are where CIKM drafts leak credibility. Terms with lane- specific meanings — entity, knowledge, relevance, scalable, online — must be pinned at first use. A table of rewrites that recur in CIKM drafts:
| Draft habit | Lane it annoys | Repair |
|---|---|---|
| "our knowledge-aware model" (never defined) | KM-DB | Name the structure: "a model conditioned on a curated product KG (2.1M triples)" |
| "significantly better" without a test | IR | Reserve the word for tested significance; else give the margin |
| "scales to large graphs" | mining | Give the size, hardware, and growth curve actually measured |
| "real-world dataset" | all three | Name the source, size, license, and collection window |
| "state-of-the-art baseline" | mining + IR | Cite venue + year; stale SOTA claims are checked at review |
test cases passed, vs 12 / 26 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.