Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/chi-related-work
Use when building the related-work section of an ACM CHI paper — covering the HCI venue family and the non-CS disciplines CHI draws on, positioning against the last few CHI cycles, keeping self-citation anonymous — in a venue where insufficient contextualization is the top assisted desk-reject ground.
Passed 22 of 23 test cases with the skill, and 17 of 23 without it.
and 3 more in the full report
5 fixed17 passed either way1 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 5 test cases the skill fixed. Same model, same prompt, with and without the skill.
In a diagnostic literature report for a CHI submission, what exact bracketed line header must precede the preprint sweep date and status assessment?
| Checked | Without skill | With skill |
|---|---|---|
| Uses the bracketed key '[Concurrent work]' for the preprint sweep line. | Missed | Passed |
The instructions your model reads when the skill loads.
At CHI, related work is screened before it is reviewed: ADR-Context — "grossly insufficient literature review to contextualize the contribution" — was the most frequently cited assisted desk-reject reason at CHI 2026, applied jointly by subcommittee chairs and ACs before external review (chi2026.acm.org outcome posts, read 2026-07-08). A thin or misaimed section no longer just lowers scores; it can end the submission in October.
| Lane | What belongs there | Who notices its absence |
|---|---|---|
| CHI itself, last ~3 cycles | The ongoing conversation your paper joins | 1AC/2AC — instantly |
| The HCI family (CSCW, UIST, DIS, IUI, ASSETS, IMWUT, TOCHI, IJHCS) | The specialist thread of your topic | The external recruited from that thread |
| Foundational HCI | The classic your framing stands on (or contradicts) | Senior reviewers |
| Feeder disciplines | Psychology, sociology, design, communication, health, ML — wherever your constructs come from | The reviewer trained in that discipline |
| Domain/application literature | The practice context you study (education, accessibility, health...) | The reviewer who lives in that context |
The feeder-discipline lane is CHI-distinctive and the most common hole: a paper measuring "trust" that cites no trust literature outside HCI, or one deploying thematic analysis citing only its HCI users, reads as unserious to exactly the reviewer the subcommittee recruited to check it.
The paragraph unit that survives review states a front — what the cited cluster established, where it stops, and how your work moves it:
Prior deployments of voice assistants with older adults established adoption barriers in home settings [x, y, z], but studied single-user households; multi-resident care contexts, where addressee ambiguity dominates, remain unexamined. We provide the first such deployment.
Three tests per paragraph: it cites ≥2 works in tension or sequence (else it is an annotation), it ends with a gap or relationship claim, and deleting it would weaken the introduction's contribution list. A related-work section whose paragraphs all pass is positioning; anything else is coverage theater — and coverage theater at excessive length also feeds the ADR-Contribution ground (small contribution relative to paper length).
test cases passed, vs 17 / 23 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.