sidiangongyuan/codex-skills-library / skills/paper-writing
paper-writing
VerifiedDraft 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.
| File | Kind | Size |
|---|---|---|
| LICENSE | other | 1.0 KB |
| SKILL.md | skill md | 7.2 KB |
| agents/openai.yaml | config | 325 B |
| references/benchmark.md | markdown | 2.8 KB |
| references/closing-material.md | markdown | 1.9 KB |
| references/consistency.md | markdown | 4.1 KB |
| references/experiments.md | markdown | 5.8 KB |
| references/front-matter.md | markdown | 4.2 KB |
| references/method.md | markdown | 1.9 KB |
| references/prose.md | markdown | 3.1 KB |
| references/related-work.md | markdown | 1.8 KB |
references/prose.md
View on GitHub# Local Prose and Transitions
Read the sentence in its paragraph and inspect the preceding and following
passages. Identify whether the problem is wording, a missing connection, a false
premise, or a category error before rewriting.
- Prefer a clear subject and action. Split nested clauses or distant coordinated
verbs when they obscure what acts on what; sentence length alone is not an error.
- Use familiar technical terms consistently rather than swapping synonyms for
variety. Define an acronym when it becomes useful, not to make prose look technical.
- Replace vague links such as "these signals" with the actual object when their
referent is ambiguous. A better noun cannot fix an undefined data relationship.
- Treat generic intensifiers, repeated contrast words, and nominalized phrases as
signals to inspect, not forbidden tokens. Keep them when they convey real meaning.
- Prefer explicit logical relations to decorative connectors. "Furthermore" does
not motivate a new topic, and "therefore" does not establish a causal inference.
- Scope necessity terms such as "requires," "only if," and "cannot" to what is
actually necessary. Explain the benefit of a design without inventing impossibility.
- Fix a repeatedly misunderstood sentence at the source instead of adding a
defensive paragraph elsewhere. Narrow an overbroad subject as well as its verb.
- Keep parallel conditions parallel; do not imply chronological recovery with
"subsequently" when two independent configurations were compared.
## Direct result prose
- End a result paragraph on the supported finding, contribution, or implication
when a complete table already shows a secondary metric or tradeoff. Do not
repeat that value as a negative rhetorical turn.
- Use “but,” “however,” and “although” when the contrast changes the logic of the
claim. Do not use them to append a self-protective caveat after a positive result.
Likewise, avoid “we acknowledge” and “it should be noted” when they only
pre-empt criticism.
- Retain a condition that changes fairness, applicability, or claim scope. Place
it where the reader interprets the comparison and state it directly, rather than
deleting it or making every result sentence end with it.
Preserve technical meaning, equations, and verified evidence. Match the project's
punctuation and notation conventions; do not impose a universal ban on a mark or
a fixed sentence budget. Punctuate displayed equations as parts of their sentences
and define symbols where readers first need them.
Write for the external reader. Internal run names, revision stages, failed launch
history, local paths, and editorial instructions do not belong in manuscript prose.
Scientifically necessary implementation details belong in the reproducibility
description; unrelated operational history need not appear even in the appendix.
After a local edit, check its claim and terminology against connected passages.
Do not use local refinement as permission to change the whole paper's positioning.
When the passage is already clear and correct, report that rather than supplying
cosmetic alternatives merely to appear productive.
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 |