Skip to content

sidiangongyuan/codex-skills-library / skills/paper-writing

paper-writing

Verified

Draft or restructure research-paper sections, refine local prose, and audit manuscript-wide consistency. Use for evidence-grounded academic writing and revision, not independent acceptance-risk review or official rebuttal management.

FileKindSize
LICENSEother1.0 KB
SKILL.mdskill md7.2 KB
agents/openai.yamlconfig325 B
references/benchmark.mdmarkdown2.8 KB
references/closing-material.mdmarkdown1.9 KB
references/consistency.mdmarkdown4.1 KB
references/experiments.mdmarkdown5.8 KB
references/front-matter.mdmarkdown4.2 KB
references/method.mdmarkdown1.9 KB
references/prose.mdmarkdown3.1 KB
references/related-work.mdmarkdown1.8 KB

references/closing-material.md

View on GitHub
# Conclusion, Limitations, and Supplement

Close on what the work enables or establishes, using the same claim boundaries as
the experiments. Do not replay every module, label type, training variant, and
control condition. A verified headline number can help when it anchors the main
takeaway; neither a numerical ending nor a number-free ending is mandatory. Do
not make a secondary cost already visible in a complete table the final sentence;
end on the contribution or supported implication.

Do not introduce new results, mechanisms, terminology, or universal claims in the
Conclusion. For a benchmark, the takeaway can concern the learning and evaluation
foundation and what using it revealed, rather than the reference model alone.
Use enough space for the actual argument without enforcing a paragraph quota.

Limitations belong where readers and the venue need them. State the affected
scope and available evidence directly; do not turn a known limitation into vague
future work to sound confident. Equally, do not repeat defensive disclaimers in
every section or restate a table-visible tradeoff as a negative closing clause.
A material limitation that changes fairness, applicability, or the central claim
must remain visible; a pointer to a limitations section is allowed when useful.

Use the supplement for extended construction details, prompts, annotation
examples, metric definitions, reproducibility settings, and additional results
that the paper actually has. Refer readers to the relevant section at the point
of need. Do not hide a condition there that changes the meaning of a main result.

Supplementary examples must distinguish supplied labels, model outputs, and
editorial illustration. Preserve real prompts and outputs as evidence; explain a
historical terminology difference rather than silently rewriting the record.
Remove obsolete descriptions and internal revision notes from reader-facing text,
not the records needed to reproduce the research.
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: 472kUpdated Claude CodeCodexVerifiedInstall
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: 472kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall