Design a literature search
Translate a research question into concepts, controlled vocabulary, text terms, database-specific queries, limits, and a documented search log.
Codex Skills for Academic Research give Codex a structured way to plan literature searches, track inclusion decisions, organize references, extract evidence, and draft a transparent synthesis. Download the skills and adapt them to your discipline, methodology, databases, and citation style.
task: map evidence on remote labs
inspect:
- research question
- eligibility criteria
- study metadata
verify: citations + decision logAcademic research depends on explicit questions, appropriate databases, documented search strategies, screening criteria, accurate metadata, critical appraisal, and careful distinction between what a study reports and what the reviewer concludes.
These skills give Codex a workflow for literature reviews and other scholarly tasks while preserving source records, eligibility decisions, evidence tables, citation details, and uncertainty.
Translate a research question into concepts, controlled vocabulary, text terms, database-specific queries, limits, and a documented search log.
Deduplicate records, apply stated eligibility criteria, record exclusion reasons, link multiple reports, and keep the study as the unit of interest.
Build consistent evidence tables for methods, samples, outcomes, limitations, and risk-of-bias judgments without inventing missing data.
Group findings by the research question, distinguish results from interpretation, preserve contradictory evidence, and keep citations attached to claims.
The workflow keeps searching, screening, extraction, appraisal, and writing as separate stages. That makes errors easier to find and the final work easier to reproduce.
Specify the question, scope, eligible study designs, population or context, outcomes, time range, languages, databases, and planned synthesis.
Run database-appropriate strategies, export full metadata, retain search dates and counts, merge results, and flag duplicates.
Apply the same criteria to each record, record reasons for exclusion, link study reports, and use a predefined extraction form.
Compare methods and findings, assess study quality, explain heterogeneity, identify gaps, and avoid claims stronger than the evidence.
A narrative review, scoping review, systematic review, and evidence map require different levels of protocol, search breadth, screening, and reporting.
Find and organize scholarship around a focused question, theoretical debate, method, or historical development.
Support protocol-driven searching, study selection, extraction, appraisal, synthesis, and transparent reporting.
Map the extent, concepts, methods, populations, and gaps in a broad or emerging body of literature.
Manage a long-running research question, reading notes, chapter evidence, citations, methods decisions, and revision checks.
The workflow can be narrowed for a course paper or expanded for a protocol-led review.
These skills are for researchers who need an organized scholarly workflow and a visible record of how evidence moved into the final document.
Structure literature searches, reading notes, evidence matrices, thesis chapters, citations, and revision checks.
Support repeatable searching, screening, extraction, appraisal, and synthesis across collaborative projects.
Follow consistent eligibility, metadata, note-taking, and evidence-extraction procedures set by the project lead.
Maintain search logs, screening decisions, evidence tables, disagreement notes, and reporting materials across the review process.
The skills assist research organization and synthesis. They do not replace subject expertise, librarian support, ethics review, statistical judgment, peer review, or verification of every citation and extracted value.
Install the skills inside one research repository or make them available across your academic projects.
Keep each SKILL.md file with included protocol checklists, search guidance, extraction templates, and reference rules.
Use a repository skill for one study or review protocol, or the user location for a reusable academic workflow.
Record the discipline, databases, eligibility criteria, citation style, reporting guideline, data fields, and prohibited assumptions.
Codex detects installed skills automatically. Restart Codex if a newly added skill is missing.
Provide the research question, review type, scope, databases, date limits, inclusion criteria, and expected output before searching.
Answers about literature reviews, citations, databases, PDFs, academic integrity, and human review.
No. They can support search planning, records, screening workflows, extraction, synthesis, and reporting, but a valid systematic review requires qualified human decisions and appropriate methodology.
Only when the environment provides authorized access. The workflow can still help prepare database-specific strategies and process exported records.
They must be verified. The skills can preserve metadata, check DOI records, and audit in-text citations, but researchers should confirm every reference against the original source.
Yes, when the files are available and readable. The workflow separates direct study findings from reviewer interpretation and records missing or unclear information.
They encourage source attribution, quotation control, citation checks, and original synthesis. Users remain responsible for following institutional rules on AI assistance and authorship.
Clear protocols. Organized records. Verified citations. Conclusions that stay close to the evidence.