Use when calibrating completeness vs. selectivity and ensuring fairness across theoretical schools, methods, authors, and debates in an Annual Review of Sociology (ARSoc) review — including not over-promoting the author's own work. Audits coverage and even-handedness; it does not design the framework (arsoc-organizing-framework) or write prose (arsoc-writing-style).
SkillScore91Maintained
Automatic
◆Research pack
Last verified 8/15/2026
What it changes
Passed 20 of 22 test cases with the skill, and 17 of 22 without it.
Without skill77%
With skill91%
Checks it passed that the model alone missed
Recommends explicitly labeling the author's position as an authorial interpretation.
Advises evaluating methodological diversity across qualitative, quantitative, and computational approaches.
Explains that comprehensive coverage is demonstrated through the citation set and saturation log rather than narrative length.
and 1 more in the full report
3 fixed17 passed either way2 still failing0 broke
Pass-rate lift
+14 pp
Tokens
+28%
Agent turns
0%
Response time
-23%
Model gemini-3.6-flashMeasured 8/15/2026v120 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 3
Prompt
In a survey on labor market polarization, two prominent computational models yield opposing predictions on future middle-skill job loss. The author prefers Model A and plans to write: 'This debate is now settled in favor of Model A.' How should live academic disagreements be handled?
What was checked, without and with the skill
Checked
Without skill
With skill
Rejects declaring an active scholarly debate resolved by author assertion.
Passed
Passed
Recommends specifying what empirical evidence would be required to settle the debate.
Passed
Passed
Recommends explicitly labeling the author's position as an authorial interpretation.
Missed
Passed
case-09 · gemini-3.6-flash · 8/15/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
The framework is set and you are filling cells with the literature
You are unsure whether to cite everything or curate
The subfield has rival theoretical schools, a live debate, or conflicting empirical findings
You are a contributor to this literature and worry the review tilts toward your own work or method
Completeness vs. selectivity: the ARSoc contract
An ARSoc review must be comprehensive in coverage yet selective in emphasis. The reader should trust that nothing important is missing, while the prose foregrounds what matters. Resolve the tension by tiering the corpus:
Tier
Treatment
Foundational / field-defining
discussed in the text, with what they established and their limits
Important contributions
grouped and weighed within the framework's cells; cited with their finding
Confirmatory / incremental
cited in clusters ("see also …") to demonstrate coverage without bloating prose
Tangential
cited only where they bear on a specific claim; not every adjacent study belongs
Comprehensiveness is proven by the citation set (the saturation log from arsoc-literature-synthesis); selectivity is exercised in the prose. A review that discusses every study at equal length has abdicated the editorial judgment that is its value.
Fairness across schools, methods, authors, and findings
ARSoc is the survey of record for a subfield: its account of a debate becomes the discipline's shared reference. That obliges genuine even-handedness:
Steelman every camp. State each theoretical school's strongest case in terms its proponents would accept before noting its weaknesses. Refute positions at their best, never at their worst.
Respect methodological pluralism. Sociology runs on quantitative, qualitative, computational, and theoretical traditions; weigh ethnographic and interpretive evidence on its own terms rather than measuring it against a regression yardstick (or vice versa). Dismissing a whole mode is a classic ARSoc balance failure.
Weigh, don't tally. Conflicting results are reconciled by credibility (design, data, transparency) and by what each studies, not by vote-counting "7 studies find positive, 4 negative."
Attribute ideas correctly. Credit originators, not just popularizers; this is a recurring referee complaint, sharpened in a discipline attentive to whose contributions get erased.
Handle live debates without resolving by fiat. Lay out the disagreement, what evidence would settle it, and where the author's own read sits — labelled as the author's read, not as consensus.
The self-citation trap
If you have published in the subfield, the review must not read as a retrospective of your program. Guardrails:
Your own work is cited at the tier its importance to the field warrants — no more.
Rival schools and rival methods to your own are given their strongest statement.
A reader who does not know who wrote the review cannot tell from the emphasis or the method mix.
When the field disagrees with you, say so plainly and represent the other side fairly.