Use when polishing the prose of an Accounting, Organizations and Society (AOS) manuscript — sustained argumentative writing that carries theory and evidence together, a self-contained abstract, disciplined length despite the absence of a hard cap, and full anonymization for double-anonymized review. Polishes prose; it does not run the submission preflight (aos-submission).
SkillScore82Maintained
Automatic
◆Writing & Editing pack
Last verified 8/15/2026
What it changes
Passed 18 of 22 test cases with the skill, and 11 of 22 without it.
Without skill50%
With skill82%
Checks it passed that the model alone missed
The response states that the abstract must contain zero citations.
The response sets the exact value under 【Next step】 to aos-submission.
The response specifies that the abstract must consist of a single self-contained paragraph.
and 7 more in the full report
7 fixed11 passed either way4 still failing0 broke
Pass-rate lift
+32 pp
Tokens
+30%
Agent turns
0%
Response time
-19%
Model gemini-3.6-flashMeasured 8/15/2026v118 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 7 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 7
Prompt
I am preparing a qualitative research manuscript on management control practices for submission to Accounting, Organizations and Society. Could you review my paper draft to make sure the writing style and structure meet the journal's expectations? Please structure your evaluation report using the standard headings: 【Register】, 【Abstract】, 【Terminology】, 【Length】, 【Anonymity & mechanics】, and 【Next step】.
What was checked, without and with the skill
Checked
Without skill
With skill
The response explicitly states that generative AI writing assistance must be declared per Elsevier policy under 【Anonymity & mechanics】.
Missed
Passed
The response sets the exact value under 【Next step】 to aos-submission.
Missed
Passed
case-01 · gemini-3.6-flash · 8/15/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md69 linessingle file
Writing Style (aos-writing-style)
When to trigger
The draft reads as description (field papers) or as a mechanical hypothesis mill (experiments)
Sections don't argue — the theory, findings, and discussion sit side by side unconnected
The abstract summarizes the topic instead of the claim
Anonymization, references, or length discipline still need a pass
The AOS register
AOS prose is essayistic but accountable: long-form argument in which every interpretive move is anchored to evidence and every theoretical term earns its keep. The reader is interdisciplinary — an organizational sociologist and a JDM psychologist may both referee — so write for the intelligent non-specialist: define tradition-specific vocabulary at first use, and never let jargon substitute for reasoning.
Argue continuously. Each section should advance the claim, not inventory content. Findings sections in field papers interleave excerpt and interpretation — the analysis is in the prose, not after it.
Voice. First person is normal and honest in interpretive work ("we came to see the budget meetings as…"); keep it analytic, not confessional. Experimental papers use the plain predictive register: theory → prediction → design → result.
Concreteness. Ground abstractions in the practice: named artifacts (the monthly pack, the risk register), scenes, and actors' words — anonymized but vivid.
Terminology ledger. Fix one name per concept (construct, practice, site pseudonym, condition label) and hold it identical across abstract, text, exhibits, and appendices.
Length without a cap
AOS publishes long papers and states no rigid word limit in its Guide for Authors (re-check the current guide — norms evolve). Treat that freedom as a discipline test: interpretive papers commonly run 12,000–15,000 words, and every one of them must be load-bearing. Cut scene-setting that does no theoretical work, second examples that prove the same point, and literature reviews of threads the paper never rejoins.
Front matter
Abstract: one self-contained paragraph — the question, the setting/design, the conceptual answer, and what it changes in the conversation. No citations, no suspense.
Keywords chosen so both accounting and social-science indexers find the paper.
Introduction executes the positioned arc (see aos-literature-positioning) and previews the conceptual answer — AOS readers expect to know the claim before the methods.
Anonymization and mechanics
Double-anonymized review: strip names, affiliations, acknowledgments, grant numbers, and self-revealing phrasing ("in our earlier study at this site") from the manuscript file; cite your own work in the third person.
Scrub document properties/metadata; keep identifying material on the separate title page only.
References in the journal's author–date style; reconcile the reference list against the current Guide for Authors rather than trusting the citation manager.
Declare any generative-AI writing assistance per Elsevier policy (a declaration in the manuscript; AI tools cannot be authors).
The paragraph audit
For a final pass, write each paragraph's job in the margin in five words or fewer ("introduces broker
concept", "handles counter-evidence on X"). Paragraphs with no statable job get cut; adjacent paragraphs
with the same job get merged. This audit is the cheapest route to AOS-grade density.