Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when a CSCW paper is accepted and enters PACMHCI production — executing minor-change conditions, deanonymizing text and artifacts, ACM journal e-rights and TAPS steps, issue assignment, and the separate track of presenting at the next CSCW conference.
.claude/skills/brycewang-stanford-cscw-camera-ready/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 62% | 0% |
Acceptance at CSCW starts two parallel tracks that most venues fuse into one: the paper becomes a journal article in PACMHCI on ACM's production schedule, and separately the authors are invited to present at the next CSCW conference. Keep the tracks distinct in your planning — the article's publication and the talk's logistics run on different clocks with different owners.
Execute the conditions first. Many acceptances arrive as Conditional Accept with Minor Changes: a named list verified before the acceptance is final. Treat the list as a contract — do the items, document where, and resist bundling in unrequested rewrites that could reopen questions.
Deanonymization is a build step, not a find-replace. Reverse every layer the submission masked:
grant numbers) — this is the text most often still missing at the deadline because it never existed in any draft.
(cscw-artifact-evaluation's release build), and verify each resolves.
promised to participants, not merely required by review — those stay. Your consent documents, not the venue, govern what gets unmasked.
ACM production mechanics. Expect journal-standard steps: e-rights/copyright forms, final source preparation through ACM's TAPS pipeline, ORCID for authors, CCS concepts, and accessibility requirements (alt text, tagged PDFs). The precise sequence, deadlines, and any open-access fee treatment for PACMHCI CSCW articles were 待核实 at the 2026-07-08 check — take them from the acceptance email and the live ACM author instructions, not from this file or from CHI-trained habits.
Issue assignment. Historically CSCW papers appeared in paired annual issues (e.g., Vol. 8: CSCW1 April 2024, CSCW2 November 2024). How rolling acceptances map to issues is part of the soft launch — confirm your article's volume/issue/article number when production tells you, then cite yourself accordingly.
pending space — confirm which conference year your paper is assigned to before booking anything. The next edition at check time: CSCW 2026, Salt Lake City, Utah, October 10-14, 2026.
待核实; verify on the conference site early, because travel budgets and visas move slower than production deadlines.
and the concept, not the pipeline; participant-voice slides need the same paraphrase discipline as the paper.
| Item | Track | Owner | Verified against | | --- | --- | --- | --- | | Minor-change conditions executed and documented | Article | Lead author | Decision letter | | Deanonymization build (text, links, acknowledgements) | Article | Lead author | Submission-era mask list | | Consent-bound masks preserved | Article | Ethics owner | Consent forms | | E-rights, TAPS, ORCID, accessibility | Article | Corresponding author | Live ACM instructions (待核实) | | Public artifact release live and cited by DOI | Article | Artifact owner | cscw-artifact-evaluation | | Conference-year assignment confirmed | Talk | Any author | Official email | | Registration/attendance rules checked | Talk | Presenting author | cscw.acm.org (待核实) |
text[Conditions] each acceptance condition → location in final text [Identity] authors/acks restored; consent-bound pseudonyms intact [Links] every URL and DOI in the final PDF resolves (run a link check) [Metadata] title/abstract/author order identical across TAPS, the PDF, and the artifact release [Two clocks] production deadline and presentation logistics tracked separately
The venue facts here (conference dates, issue anchors) were verified 2026-07-08; production-mechanics items are flagged 待核实 above because ACM instructions arrive with the acceptance and override anything written down in advance.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→fail | 27,513 | 19,945 | -28% | 1 | 1 | 0% | 3,323 | 3,861 | +16% | 0 | 0 | — |
case-04 | pass→pass | 15,640 | 16,344 | +5% | 1 | 1 | 0% | 1,708 | 2,774 | +62% | 0 | 0 | — |
case-05 | pass→pass | 14,337 | 14,060 | -2% | 1 | 1 | 0% | 2,032 | 2,758 | +36% | 0 | 0 | — |
case-01 | fail→pass | 31,110 | 25,359 | -18% | 1 | 1 | 0% | 3,632 | 4,334 | +19% | 0 | 0 | — |
case-02 | fail→fail | 24,981 | 12,774 | -49% | 1 | 1 | 0% | 3,298 | 2,959 | -10% | 0 | 0 | — |
case-06 | pass→pass | 27,837 | 21,409 | -23% | 1 | 1 | 0% | 3,299 | 3,792 | +15% | 0 | 0 | — |
case-07 | pass→pass | 21,530 | 19,262 | -11% | 1 | 1 | 0% | 2,557 | 3,263 | +28% | 0 | 0 | — |
case-08 | pass→pass | 15,869 | 13,501 | -15% | 1 | 1 | 0% | 1,460 | 2,216 | +52% | 0 | 0 | — |
case-09 | pass→pass | 13,024 | 12,034 | -8% | 1 | 1 | 0% | 1,817 | 2,214 | +22% | 0 | 0 | — |
case-10 | fail→pass | 13,496 | 14,307 | +6% | 1 | 1 | 0% | 2,145 | 2,720 | +27% | 0 | 0 | — |
case-11 | pass→pass | 13,559 | 10,434 | -23% | 1 | 1 | 0% | 1,354 | 2,042 | +51% | 0 | 0 | — |
case-12 | pass→pass | 17,786 | 13,471 | -24% | 1 | 1 | 0% | 1,871 | 2,333 | +25% | 0 | 0 | — |
case-13 | fail→pass | 17,756 | 14,724 | -17% | 1 | 1 | 0% | 1,854 | 2,548 | +37% | 0 | 0 | — |
case-14 | pass→pass | 18,807 | 11,933 | -37% | 1 | 1 | 0% | 1,852 | 2,121 | +15% | 0 | 0 | — |
case-15 | pass→pass | 15,888 | 7,355 | -54% | 1 | 1 | 0% | 2,098 | 2,250 | +7% | 0 | 0 | — |
case-16 | pass→pass | 13,673 | 9,781 | -28% | 1 | 1 | 0% | 1,455 | 2,681 | +84% | 0 | 0 | — |
case-17 | fail→pass | 14,696 | 4,757 | -68% | 1 | 1 | 0% | 1,568 | 1,720 | +10% | 0 | 0 | — |
case-18 | pass→pass | 19,933 | 16,936 | -15% | 1 | 1 | 0% | 2,137 | 3,073 | +44% | 0 | 0 | — |
case-19 | pass→pass | 16,915 | 10,652 | -37% | 1 | 1 | 0% | 2,016 | 2,004 | -1% | 0 | 0 | — |
case-20 | pass→pass | 13,410 | 9,903 | -26% | 1 | 1 | 0% | 990 | 1,680 | +70% | 0 | 0 | — |
case-21 | pass→pass | 10,419 | 3,661 | -65% | 1 | 1 | 0% | 1,646 | 1,713 | +4% | 0 | 0 | — |
case-22 | pass→pass | 15,183 | 9,077 | -40% | 1 | 1 | 0% | 1,558 | 1,710 | +10% | 0 | 0 | — |
case-23 | pass→pass | 13,366 | 17,502 | +31% | 1 | 1 | 0% | 2,017 | 2,880 | +43% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 23 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.