sidiangongyuan/codex-skills-library / skills/research-evidence
research-evidence
VerifiedUse when research writing, reviews, rebuttals, or related-work planning needs academic evidence. Searches literature, screens candidate papers, audits novelty and missing work, verifies DOI/BibTeX/reference metadata and citation authenticity, and checks whether sources support the stated claims.
| File | Kind | Size |
|---|---|---|
| LICENSE | other | 1.0 KB |
| SKILL.md | skill md | 5.0 KB |
| agents/openai.yaml | config | 277 B |
| references/recent-literature-audit.md | markdown | 3.3 KB |
| references/tooling.md | markdown | 2.9 KB |
| scripts/arxiv_query_matrix.py | script | 8.3 KB |
references/recent-literature-audit.md
View on GitHub# Recent Literature / Novelty Risk Audit
Use this checklist when a paper's novelty, related-work coverage, reviewer risk,
or first/SOTA/new-benchmark positioning depends on not missing recent or adjacent
work.
## Claim Axes
Extract the paper claim into searchable axes before querying:
- Domain/input: the data source, sensing setup, modality, system context, or user population.
- Method family: model class, algorithm route, training style, or analysis framework.
- Task/output: prediction target, interaction format, benchmark task, or evaluation endpoint.
- Evaluation protocol: metrics, controls, robustness checks, calibration, human study, or audit style.
- Dataset/benchmark: named datasets, new benchmark type, split, annotation, or release claim.
- Reliability/failure mode: hallucination, leakage, grounding, bias, uncertainty, safety, or metric failure.
## Query Matrix
Build at least these query families for high-stakes novelty checks:
- Direct claim query: combine the strongest domain, method, and task terms.
- Method-route query: search method family plus domain, without the exact task term.
- Task/benchmark query: search the task or benchmark type plus evaluation protocol terms.
- Dataset/domain query: search dataset, sensing setup, or domain terms plus broader task words.
- Neighboring terminology query: search older names, adjacent community terms, acronyms, and alternate spellings.
- Newest broad query: search one broad domain/task expression sorted by recent submission date.
Avoid relying on one exact phrase. If the user supplies a term, search its
synonyms and adjacent expressions before concluding that no close work exists.
## Evidence Routes
Use at least two routes for high-stakes checks:
- Direct metadata search: arXiv API, official venue/open-access pages, DOI/Crossref where appropriate.
- Local sources: provided PDFs, `reference/` folders, BibTeX, LaTeX source, related-work notes, or prior paper lists.
- Wrapper search: local `paper-search` or similar tools as a smoke/cross-check, not as sole evidence.
- Citation graph or venue pages when direct metadata is inconclusive and the deadline risk is high.
State source limits. Absence of hits is not strong evidence unless query
families, dates, and source routes are explicit.
## Classification Rubric
- Direct competitor: substantially same input/domain, task/output, and evaluation claim.
- Claim limiter: adjacent work that weakens first/SOTA/only/new-benchmark language but does not replace the contribution.
- Table candidate: close enough for a benchmark or method comparison row.
- Prose citation: relevant adjacent work that belongs in related work but not in the main comparison table.
- Background: useful context but not a threat to the central claim.
- Irrelevant: mismatched task, modality, domain, or contribution after abstract/source inspection.
## Required Output
For a novelty-risk audit, report:
- Search matrix summary: query families, source routes, and date/venue scope.
- Candidate table: title, year/date, identifier, classification, and why it matters.
- Claim-impact verdict: keep, weaken, add citation/table row, revise positioning, or avoid first/SOTA language.
- Recommended paper edits: concrete citation, wording, table, or limitation changes.
- Residual risk: what may still be missing and how serious that uncertainty is.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |