Use when documenting the literature search and making any embedded meta-analysis reproducible for an Annual Review of Psychology (ARPsych) review. Covers search transparency, meta-analytic rigor, and open materials; it does not run the narrative search (arpsych-literature-synthesis) or design exhibits (arpsych-tables-figures).
SkillScore100Maintained
Automatic
◆Research pack
Last verified 8/15/2026
What it changes
Passed 22 of 22 test cases with the skill, and 18 of 22 without it.
Without skill82%
With skill100%
Checks it passed that the model alone missed
States that AI tools must not be listed as authors
Classifies the manuscript review type as an embedded meta-analysis
Identifies that a stopping rule is required in the search protocol disclosure
and 1 more in the full report
4 fixed18 passed either way0 still failing0 broke
Pass-rate lift
+18 pp
Tokens
+18%
Agent turns
0%
Response time
-38%
Model gemini-3.6-flashMeasured 8/15/2026v122 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 4
Prompt
We are preparing a pure narrative review on social anxiety mechanisms for ARPsych. We didn't perform a meta-analysis, but we documented our database search terms, date ranges, and selection criteria. Please review our reproducibility setup and output your assessment using the standard format: Review type, Search transparency reproducibility, If meta-analysis requirements, Open materials status, Heterogeneity / bias assessments, Declarations (funding/COI/AI), and Next step guidance.
What was checked, without and with the skill
Checked
Without skill
With skill
Identifies the review type as a narrative review
Passed
Passed
Directs workflow to arpsych-editor-strategy for coordinating manuscript scope and timeline with the editor
Missed
Passed
case-02 · 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 review documents a systematic search and you must report it reproducibly
The review embeds a meta-analysis or any new quantitative synthesis
You are deciding what to deposit (search log, coding sheet, effect-size data, code)
A reader or the Committee should be able to verify how the literature was selected
A review reports no new data — so transparency bites elsewhere
A pure narrative review has no dataset of its own, so the transparency obligation does not look like a primary-paper replication package. It bites on two things:
How the literature was found and selected — the search protocol from arpsych-literature-synthesis, written up so a reader could reproduce the coverage.
Any quantitative synthesis the review itself contributes — if you compute pooled effects, that is original analysis, and it must be reproducible (检索于 2026-06；以官网为准).
Post-replication-crisis, ARPsych readers expect both, and a review that asserts "the literature shows…" with no documented basis reads as less authoritative.
If the review is narrative (no meta-analysis)
Report the search: databases, terms, date range, inclusion/exclusion, and the stopping rule — a short, near-PRISMA-style account suffices.
State selection logic: why these studies and not others (especially when the field is large and you are selective).
Be explicit about replication status of contested effects (this is part of transparency, not just balance).
If the review embeds a meta-analysis
Then you have run original analysis and must meet quantitative-synthesis standards:
Requirement
What to provide
PRISMA-style flow
search → screening → included, with counts at each step
Coding protocol
how effects were extracted/coded; inter-coder reliability
Effect-size dataset
the extracted effects + moderators, deposited
Analysis code
scripts reproducing the pooled estimates and plots