Use when handling data documentation, sharing, and spatial-data ethics for an Annals of the American Association of Geographers manuscript — provenance (sources, projections, processing), reproducibility, and geoprivacy / human-subjects protection. Prepares documentation; it does not over-state a verification gate the policy does not impose.
SkillScore86Maintained
Automatic
◆Research pack
Last verified 8/15/2026
What it changes
Passed 19 of 22 test cases with the skill, and 16 of 22 without it.
Without skill73%
With skill86%
Checks it passed that the model alone missed
Includes the bracketed tag 【Provenance】
Includes the bracketed tag 【Geoprivacy】
Includes the bracketed tag 【Policy check】
and 8 more in the full report
6 fixed13 passed either way0 still failing3 broke
Pass-rate lift
+14 pp
Tokens
+21%
Agent turns
0%
Response time
-30%
Model gemini-3.6-flashMeasured 8/15/2026v119 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 6
Prompt
I am finalizing my manuscript for the Annals of the American Association of Geographers. It combines household field GPS points with public census data and spatial joins. Could you evaluate my spatial data package, privacy protections, and code documentation? Please present your evaluation using bracketed headers covering Provenance, Geoprivacy, Human subjects, Reproducibility, Data-availability statement details, Policy check confirmation, and Next step.
What was checked, without and with the skill
Checked
Without skill
With skill
Includes the bracketed tag 【Provenance】
Missed
Passed
Includes the bracketed tag 【Geoprivacy】
Missed
Passed
Includes the routing tag 【Next】aaag-review-process
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.md88 linessingle file
Transparency & Spatial Data (aaag-transparency-and-data)
The Annals (Taylor & Francis, for the AAG) follows publisher and disciplinary norms for data
availability and research integrity rather than (as at some journals) a mandatory, editor-run
replication check before publication. The right posture is: document spatial-data provenance
thoroughly, share what you ethically can, and protect location privacy. Confirm the current data-
availability / sharing policy on the Taylor & Francis journal page (检索于 2026-06；以官网为准).
When to trigger
Preparing data documentation, a data-availability statement, or supplementary materials
Deciding what spatial data can be shared given geoprivacy, IRB, license, or proprietary limits
Documenting qualitative or restricted location data so claims remain credible
A reviewer asked how others could verify or reproduce the spatial analysis
Spatial-data provenance (the geography-specific core)
Document enough that another geographer could locate, re-project, and rebuild the dataset:
Sources & licenses for every layer (census, satellite, OSM, administrative, field-collected),
with access dates and terms.
Projection / CRS and any re-projection; areal units and how they were defined or aggregated.
Processing chain — classification, geocoding, spatial joins, resampling, filtering — as a logged,
ordered pipeline, not "cleaned in GIS."
Temporal alignment — vintages of layers and how they were harmonized.
Geoprivacy & human subjects (overrides sharing)
Do not publish precise locations of identifiable individuals/households; use geographic masking
(aggregation, jittering, areal reporting) and state the method.
IRB/consent for interviews, fieldwork, and participatory mapping; protect informant and site identity.
Sensitive sites (sacred, endangered-species, conflict) may require deliberate spatial coarsening — say so.
Reproducibility (good practice even without a pre-publication gate)
Master script regenerates every table, figure, and map from raw/constructed data.
README/codebook: provenance, projections, construction steps, and how to reproduce each exhibit.
Seeds set for stochastic steps; software/package versions pinned (renv.lock / requirements.txt).
Restricted data: a data-availability statement explaining what is shared, what is not and why,
and an access path; provide synthetic/masked data so code can run where feasible.