fair-check

Name: fair-check
Author: scdenney

34stars0forks

How to install fair-check?

scdenney/open-science-skills/fair-check

$npx -y skills add scdenney/open-science-skills --skill fair-check

Installs into the current project.

Summary

Audit manuscript and replication package against FAIR open-science principles.

Files · 1

View on GitHub

SKILL.md

1	# FAIR Manuscript Checker
2
3	## Standards anchor
4
5	Use the FAIR principles as manuscript-facing checks for research objects: Findable, Accessible, Interoperable, Reusable. FAIR does not mean everything must be openly downloadable. Sensitive or restricted data can be FAIR when metadata, access conditions, identifiers, and reuse terms are explicit. The practical standard is "as open as possible, as restricted as necessary."
6
7	Core references: Wilkinson et al. (2016) for the FAIR principles, GO FAIR for the F/A/I/R subprinciples, OSF documentation for repository metadata and data archiving, FORCE11 for data citation principles, and TOP/DA-RT for manuscript transparency expectations.
8
9	## Instructions
10
11	### 1. Inventory research objects
12
13	Before judging compliance, list every research object the manuscript depends on:
14
15	- Raw data, cleaned data, derived data, and analysis-ready data.
16	- Analysis code, simulation code, randomization scripts, package lockfiles, and notebooks.
17	- Survey instruments, treatments, vignettes, questionnaires, prompts, codebooks, classification labels, OCR prompts, and annotation guidelines.
18	- Figures, tables, model outputs, topic models, classifiers, trained weights, dictionaries, and corpora.
19	- Preregistrations, PAPs, IRB/ethics protocols, consent language, and data-use agreements.
20	- Third-party data or proprietary inputs that cannot be redistributed.
21
22	If an object is not shareable, it still needs metadata and a clear access or non-availability explanation.
23
24	### 2. Check Findable
25
26	For each research object, verify:
27
28	- Repository or landing page exists and is public or publicly discoverable.
29	- Persistent identifier exists or is planned: DOI, ARK, Handle, OSF registration DOI, Zenodo DOI, Dataverse DOI, ICPSR study ID, or equivalent.
30	- Metadata includes title, creators/contributors, description, date/version, resource type, keywords/tags, funder when relevant, related publication DOI, and related object links.
31	- Data/code/materials are cited in the manuscript or reference list, not only mentioned in prose.
32	- File names are interpretable and map to manuscript tables, figures, or analyses.
33
34	Prompt author if missing: repository URL, DOI/identifier, title, contributors, version/date, and how each object maps to manuscript claims.
35
36	### 3. Check Accessible
37
38	Verify:
39
40	- Access route is clear: open download, embargoed release date, controlled access, data-use agreement, request email/form, or legal/ethical non-availability.
41	- Protocol is standard and durable: repository landing page, HTTPS, institutional repository, OSF, Dataverse, Zenodo, ICPSR, Dryad, GitHub+Zenodo, or discipline repository.
42	- Restricted data have public metadata and explicit criteria for access.
43	- Sensitive data are not uploaded to unsuitable repositories; OSF should not be used for sensitive or personally identifiable health information.
44	- Availability statements match actual repository state.
45	- Embargoes and anonymous peer-review links are handled without breaking post-publication access.
46
47	Prompt author if missing: access restrictions, embargo date, contact process, data-use agreement, privacy constraints, and post-acceptance public URL.
48
49	### 4. Check Interoperable
50
51	Verify that others can read and combine the materials:
52
53	- Data use non-proprietary or widely readable formats where possible: CSV/TSV, TXT, JSON, Parquet, RDS with fallback, Stata with codebook, PDF/A for documents.
54	- Variables, labels, missing values, scales, treatment arms, weights, and derived variables are documented.
55	- Code has an environment file or setup notes: `renv.lock`, `requirements.txt`, `environment.yml`, Dockerfile, session info, package versions, or OS notes.
56	- Metadata follows a discipline-appropriate standard where available: DDI for social-science survey data, DataCite metadata, repository community schemas, or structured README.
57	- Cross-links connect raw data, cleaned data, code, figures/tables, and manuscript claims.
58
59	Prompt author if missing: codebook, README, variable dictionary, software environment, data provenance, or mapping from files to outputs.
60
61	### 5. Check Reusable
62
63	Verify:
64
65	- License is explicit for each shareable object: data, code, text/materials, and figures may need different licenses.
66	- Reuse conditions are clear for third-party, proprietary, copyrighted, or restricted materials.
67	- Provenance is documented: collection method, source, cleaning steps, transformations, exclusions, and version history.
68	- Minimal replication path is documented: what script to run, in what order, and what output it produces.
69	- Ethics and consent allow the claimed sharing level.
70	- Data/code/material citations give credit to creators and make reuse citable.
71
72	Prompt author if missing: license choices, consent/sharing compatibility, restrictions on reuse, provenance notes, and replication instructions.
73
74	### 6. Audit manuscript statements

People who liked this also viewed

6 picks

Type

find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 20262.6M27k
grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026633k182k
vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026573k29k
improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026522k182k
setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026454k182k
triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026424k182k

fair-check

34stars0forks

View on GitHub

How to install fair-check?

scdenney/open-science-skills/fair-check

$npx -y skills add scdenney/open-science-skills --skill fair-check

Installs into the current project.

Summary

Audit manuscript and replication package against FAIR open-science principles.

Files · 1

View on GitHub

SKILL.md

1	# FAIR Manuscript Checker
2
3	## Standards anchor
4
5	Use the FAIR principles as manuscript-facing checks for research objects: Findable, Accessible, Interoperable, Reusable. FAIR does not mean everything must be openly downloadable. Sensitive or restricted data can be FAIR when metadata, access conditions, identifiers, and reuse terms are explicit. The practical standard is "as open as possible, as restricted as necessary."
6
7	Core references: Wilkinson et al. (2016) for the FAIR principles, GO FAIR for the F/A/I/R subprinciples, OSF documentation for repository metadata and data archiving, FORCE11 for data citation principles, and TOP/DA-RT for manuscript transparency expectations.
8
9	## Instructions
10
11	### 1. Inventory research objects
12
13	Before judging compliance, list every research object the manuscript depends on:
14
15	- Raw data, cleaned data, derived data, and analysis-ready data.
16	- Analysis code, simulation code, randomization scripts, package lockfiles, and notebooks.
17	- Survey instruments, treatments, vignettes, questionnaires, prompts, codebooks, classification labels, OCR prompts, and annotation guidelines.
18	- Figures, tables, model outputs, topic models, classifiers, trained weights, dictionaries, and corpora.
19	- Preregistrations, PAPs, IRB/ethics protocols, consent language, and data-use agreements.
20	- Third-party data or proprietary inputs that cannot be redistributed.
21
22	If an object is not shareable, it still needs metadata and a clear access or non-availability explanation.
23
24	### 2. Check Findable
25
26	For each research object, verify:
27
28	- Repository or landing page exists and is public or publicly discoverable.
29	- Persistent identifier exists or is planned: DOI, ARK, Handle, OSF registration DOI, Zenodo DOI, Dataverse DOI, ICPSR study ID, or equivalent.
30	- Metadata includes title, creators/contributors, description, date/version, resource type, keywords/tags, funder when relevant, related publication DOI, and related object links.
31	- Data/code/materials are cited in the manuscript or reference list, not only mentioned in prose.
32	- File names are interpretable and map to manuscript tables, figures, or analyses.
33
34	Prompt author if missing: repository URL, DOI/identifier, title, contributors, version/date, and how each object maps to manuscript claims.
35
36	### 3. Check Accessible
37
38	Verify:
39
40	- Access route is clear: open download, embargoed release date, controlled access, data-use agreement, request email/form, or legal/ethical non-availability.
41	- Protocol is standard and durable: repository landing page, HTTPS, institutional repository, OSF, Dataverse, Zenodo, ICPSR, Dryad, GitHub+Zenodo, or discipline repository.
42	- Restricted data have public metadata and explicit criteria for access.
43	- Sensitive data are not uploaded to unsuitable repositories; OSF should not be used for sensitive or personally identifiable health information.
44	- Availability statements match actual repository state.
45	- Embargoes and anonymous peer-review links are handled without breaking post-publication access.
46
47	Prompt author if missing: access restrictions, embargo date, contact process, data-use agreement, privacy constraints, and post-acceptance public URL.
48
49	### 4. Check Interoperable
50
51	Verify that others can read and combine the materials:
52
53	- Data use non-proprietary or widely readable formats where possible: CSV/TSV, TXT, JSON, Parquet, RDS with fallback, Stata with codebook, PDF/A for documents.
54	- Variables, labels, missing values, scales, treatment arms, weights, and derived variables are documented.
55	- Code has an environment file or setup notes: `renv.lock`, `requirements.txt`, `environment.yml`, Dockerfile, session info, package versions, or OS notes.
56	- Metadata follows a discipline-appropriate standard where available: DDI for social-science survey data, DataCite metadata, repository community schemas, or structured README.
57	- Cross-links connect raw data, cleaned data, code, figures/tables, and manuscript claims.
58
59	Prompt author if missing: codebook, README, variable dictionary, software environment, data provenance, or mapping from files to outputs.
60
61	### 5. Check Reusable
62
63	Verify:
64
65	- License is explicit for each shareable object: data, code, text/materials, and figures may need different licenses.
66	- Reuse conditions are clear for third-party, proprietary, copyrighted, or restricted materials.
67	- Provenance is documented: collection method, source, cleaning steps, transformations, exclusions, and version history.
68	- Minimal replication path is documented: what script to run, in what order, and what output it produces.
69	- Ethics and consent allow the claimed sharing level.
70	- Data/code/material citations give credit to creators and make reuse citable.
71
72	Prompt author if missing: license choices, consent/sharing compatibility, restrictions on reuse, provenance notes, and replication instructions.
73
74	### 6. Audit manuscript statements

People who liked this also viewed

6 picks

Type