Use when systematically gathering, reading, and structuring the literature for an Annual Review of Psychology (ARPsych) review so coverage is comprehensive and current. Builds the evidence base; it does not design the analytical spine (arpsych-organizing-framework) or judge balance (arpsych-comprehensiveness-and-balance).
SkillScore82Maintained
Automatic
◆Research pack
Last verified 8/15/2026
What it changes
Passed 18 of 22 test cases with the skill, and 7 of 22 without it.
Without skill32%
With skill82%
Checks it passed that the model alone missed
The output contains the tag 【Coverage】
The output flags Meta-analysis planned? as Y
The output includes the tag 【Search protocol】
and 28 more in the full report
15 fixed3 passed either way0 still failing4 broke
Pass-rate lift
+50 pp
Tokens
+14%
Agent turns
0%
Response time
-25%
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 15 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 15
Prompt
I have completed a saturated literature search on metacognition in educational psychology with a full matrix tracking design, replication, and lab concentration. Report the assessment summary and specify where I should route next.
What was checked, without and with the skill
Checked
Without skill
With skill
The next step explicitly directs to arpsych-organizing-framework
Missed
Passed
The output uses the exact bracketed header format 【Next step】
Missed
Passed
case-13 · 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.md71 linessingle file
Literature Synthesis — Reading the Whole Field (arpsych-literature-synthesis)
When to trigger
An invitation is in hand and the reading must start in earnest
The user has "read a lot" but has no system for tracking what is covered
Coverage gaps are likely (one paradigm, one lab, or one decade dominates the notes)
An embedded meta-analysis is planned and the search must be documentable
Coverage is the price of admission, not the contribution
ARPsych readers trust a review to be comprehensive within its frame and current to the frontier. A reviewer caught missing a major line of work loses authority instantly. But coverage is the floor: the contribution is the framework you impose later (arpsych-organizing-framework). This skill builds an evidence base broad and well-recorded enough that (a) nothing important is missed and (b) you can later be selective in prose without being incomplete in coverage.
A systematic, documentable search
Even a narrative review benefits from a near-systematic search you can describe:
Seed set — the canonical and recent agenda-setting papers you already know.
Forward + backward citation chasing — who cites the seeds; whom the seeds cite.
Database sweep — PsycINFO/APA PsycNet, Web of Science, Scopus, Google Scholar with explicit terms.
Specialty + adjacent venues — the relevant primary journals AND adjacent fields (neuro, economics, linguistics) the topic touches.
Stopping rule — saturation: new searches surface only papers already in the matrix.
Record the protocol — terms, databases, dates, inclusion/exclusion — even for a narrative review; mandatory if you report a meta-analysis (→ arpsych-transparency-and-reproducibility).
Document the search as you go, not from memory afterward. ARPsych's transparency expectations and any meta-analytic component both depend on a search log you can reproduce (检索于 2026-06；以官网为准).
The evidence matrix (one row per study/finding)
Build a structured table — not a pile of PDFs. Useful columns:
Column
Why
Construct / question
lets you later sort papers into framework cells
Population / paradigm
surfaces over-reliance on one method or sample (WEIRD bias)
Design + strength
feeds credibility weighting in arpsych-comprehensiveness-and-balance
Effect / finding (+ direction, size if available)
enables quantitative synthesis
Replication status
flags shaky vs. robust effects (post-replication-crisis hygiene)
Lab / camp
exposes citation concentration and balance risks
The matrix is the bridge: its rows become framework cells, its columns become balance and credibility checks.
Use when systematically gathering, coding, and synthesizing a management/organization literature for an Academy of Management Annals (Annals) review — the search-and-coverage methodology and the choice of narrative vs. systematic vs. bibliometric integration. Builds the corpus and synthesis notes; it does not impose the organizing spine (amann-organizing-framework) or write prose.