Nature-Style Scientific Writing
Use this skill when the user needs help creating or rebuilding manuscript prose,
not merely polishing existing sentences.
Core stance
- Author evidence comes first. Do not invent results, mechanisms, references,
methods, novelty, sample sizes, statistics or limitations.
- Write the argument before writing the sentences.
- Make the paper easy to judge: relevance, novelty, trust, reuse and meaning.
- Use ambitious but bounded claims.
- If essential evidence is missing, write a placeholder or ask for the missing
input instead of filling the gap.
When to open extra files
| File |
Open when |
| references/article-architecture.md |
You need section-level structure, argument order, or published-article writing patterns |
| references/abstract.md |
Drafting or revising an abstract, especially challenge-contribution and challenge-insight-contribution forms |
| references/introduction.md |
Drafting or revising an Introduction, task framing, technical challenge, contribution framing, or teaser/pipeline logic |
| references/related-work.md |
Rebuilding Related Work as topic synthesis instead of a paper-by-paper list |
| references/method.md |
Writing Method sections, pipeline modules, module motivation, technical advantages, or implementation details |
| references/experiments.md |
Planning or writing Experiments/Results around baselines, ablations, metrics, tables, figures, and claim support |
| references/conclusion.md |
Writing a bounded conclusion with contribution, evidence, impact, limitation, and future direction |
| references/paragraph-flow.md |
User asks whether a paragraph flows, makes sense, or is clear; use reverse outlining and paragraph-message checks |
| references/paper-review.md |
Final manuscript self-review, rejection-risk audit, claim-evidence alignment, or reviewer-facing critique |
| references/chinese-author-workflow.md |
The user's notes are Chinese, mixed Chinese-English, or organized as lab notes rather than manuscript prose |
| references/examples/index.md |
You need concrete abstract, introduction, or method examples after choosing the relevant guide |
Intake
Before drafting, identify:
- manuscript section: title, abstract, introduction, results, discussion,
conclusion, significance paragraph or full outline
- paper type: mechanism, method, resource, device, model, clinical, materials,
computational or interdisciplinary
- core claim: what the paper actually demonstrates
- evidence: figures, measurements, comparisons, datasets, statistics or examples
- boundary: where the claim stops
- target journal or word limit, if provided
If any of core claim, evidence or boundary is absent, expose the gap before
drafting. You may still produce a scaffold with explicit placeholders.
Writing workflow
- Build a one-sentence argument:
In [system/problem], we show [advance] using [approach], supported by [evidence], with [boundary].
- Choose the section architecture from
references/article-architecture.md.
- Map each paragraph to one job: context, gap, approach, result, comparison,
mechanism, implication or limitation.
- Draft from evidence outward. Keep claims near the data that support them.
- Calibrate verbs:
show, demonstrate, suggest, indicate, enable,
may, could.
- Remove unsupported novelty and universal claims.
- Run a paragraph-flow check: one paragraph, one message, with a clear first
sentence and explicit sentence-to-sentence relation.
- Return prose plus concise notes on assumptions and missing inputs.
Section defaults
Abstract
Default Nature pattern:
context/problem -> gap -> approach -> key result -> implication -> boundary
For technical AI, ML, CV or method-heavy manuscripts, open
references/abstract.md and choose one of:
challenge -> contributionchallenge -> insight -> contributionmultiple contributions
Keep it compact. Include quantitative or comparative detail when the user
provided it. End with what the work enables, not generic importance.
Introduction
Use:
field scale -> bottleneck -> prior attempts -> unresolved gap -> present study
For method-heavy papers, open references/introduction.md and reason backward
from the technical challenge and contribution before drafting forward.
Do not summarize all results. The final paragraph should state what this paper
does and how it addresses the gap.
Results narrative
Use an evidence ladder:
system/workflow -> validation -> main result -> baseline comparison -> mechanism/diagnostic analysis -> application or generalization
Each subsection should have a claim-first opening and then data support.
For ML/conference-style experiment sections, open references/experiments.md
and make sure each major claim is backed by comparison, ablation, or stress-test
evidence.
Related Work
Use:
topic scope -> representative methods -> limitation tied to this paper -> distinction
Group prior work by technical topic and mechanism, not by publication year.
Discussion
Use:
central advance -> evidence meaning -> relation to prior work -> constraints -> future use
This is where interpretation and limitations belong. Do not repeat the Results
section figure by figure.
Conclusion
Use:
contribution -> decisive evidence -> implication -> boundary
No new data. No unsupported promises.
Title
Prefer concrete titles that combine:
system/object + action/capability + application or consequence
Avoid slogan titles, grant-style aims and overbroad field claims.
Output format
Default output:
Draft: with the requested prose.Section outline: with 3-7 compact bullets when the task involves a full section.Assumptions or missing inputs: with only material issues.Claim-evidence map: for major claims, using Claim: ... | Evidence: ... | Status: supported/needs evidence.Why this structure: with 2-4 short bullets.
For Chinese author notes, provide polished English first, then brief Chinese
notes explaining major structural choices.
1---2name: nature-writing3description: Draft, restructure, or plan Nature-style manuscript sections from author-provided claims, results, figures, notes, or Chinese drafts. Use when the user wants to write or rebuild an abstract, introduction, results narrative, discussion, conclusion, title, or full manuscript argument rather than only polish finished prose.4---5
6# Nature-Style Scientific Writing
7
8Use this skill when the user needs help creating or rebuilding manuscript prose,
9not merely polishing existing sentences.
10
11## Core stance
12
13- Author evidence comes first. Do not invent results, mechanisms, references,
14 methods, novelty, sample sizes, statistics or limitations.
15- Write the argument before writing the sentences.
16- Make the paper easy to judge: relevance, novelty, trust, reuse and meaning.
17- Use ambitious but bounded claims.
18- If essential evidence is missing, write a placeholder or ask for the missing
19 input instead of filling the gap.
20
21## When to open extra files
22
23| File | Open when |
24|---|---|
25| [references/article-architecture.md](references/article-architecture.md) | You need section-level structure, argument order, or published-article writing patterns |
26| [references/abstract.md](references/abstract.md) | Drafting or revising an abstract, especially challenge-contribution and challenge-insight-contribution forms |
27| [references/introduction.md](references/introduction.md) | Drafting or revising an Introduction, task framing, technical challenge, contribution framing, or teaser/pipeline logic |
28| [references/related-work.md](references/related-work.md) | Rebuilding Related Work as topic synthesis instead of a paper-by-paper list |
29| [references/method.md](references/method.md) | Writing Method sections, pipeline modules, module motivation, technical advantages, or implementation details |
30| [references/experiments.md](references/experiments.md) | Planning or writing Experiments/Results around baselines, ablations, metrics, tables, figures, and claim support |
31| [references/conclusion.md](references/conclusion.md) | Writing a bounded conclusion with contribution, evidence, impact, limitation, and future direction |
32| [references/paragraph-flow.md](references/paragraph-flow.md) | User asks whether a paragraph flows, makes sense, or is clear; use reverse outlining and paragraph-message checks |
33| [references/paper-review.md](references/paper-review.md) | Final manuscript self-review, rejection-risk audit, claim-evidence alignment, or reviewer-facing critique |
34| [references/chinese-author-workflow.md](references/chinese-author-workflow.md) | The user's notes are Chinese, mixed Chinese-English, or organized as lab notes rather than manuscript prose |
35| [references/examples/index.md](references/examples/index.md) | You need concrete abstract, introduction, or method examples after choosing the relevant guide |
36
37## Intake
38
39Before drafting, identify:
40
41- manuscript section: title, abstract, introduction, results, discussion,
42 conclusion, significance paragraph or full outline
43- paper type: mechanism, method, resource, device, model, clinical, materials,
44 computational or interdisciplinary
45- core claim: what the paper actually demonstrates
46- evidence: figures, measurements, comparisons, datasets, statistics or examples
47- boundary: where the claim stops
48- target journal or word limit, if provided
49
50If any of `core claim`, `evidence` or `boundary` is absent, expose the gap before
51drafting. You may still produce a scaffold with explicit placeholders.
52
53## Writing workflow
54
551. Build a one-sentence argument: `In [system/problem], we show [advance] using
56 [approach], supported by [evidence], with [boundary].`
572. Choose the section architecture from `references/article-architecture.md`.
583. Map each paragraph to one job: context, gap, approach, result, comparison,
59 mechanism, implication or limitation.
604. Draft from evidence outward. Keep claims near the data that support them.
615. Calibrate verbs: `show`, `demonstrate`, `suggest`, `indicate`, `enable`,
62 `may`, `could`.
636. Remove unsupported novelty and universal claims.
647. Run a paragraph-flow check: one paragraph, one message, with a clear first
65 sentence and explicit sentence-to-sentence relation.
668. Return prose plus concise notes on assumptions and missing inputs.
67
68## Section defaults
69
70### Abstract
71
72Default Nature pattern:
73
74`context/problem -> gap -> approach -> key result -> implication -> boundary`
75
76For technical AI, ML, CV or method-heavy manuscripts, open
77`references/abstract.md` and choose one of:
78
79- `challenge -> contribution`
80- `challenge -> insight -> contribution`
81- `multiple contributions`
82
83Keep it compact. Include quantitative or comparative detail when the user
84provided it. End with what the work enables, not generic importance.
85
86### Introduction
87
88Use:
89
90`field scale -> bottleneck -> prior attempts -> unresolved gap -> present study`
91
92For method-heavy papers, open `references/introduction.md` and reason backward
93from the technical challenge and contribution before drafting forward.
94
95Do not summarize all results. The final paragraph should state what this paper
96does and how it addresses the gap.
97
98### Results narrative
99
100Use an evidence ladder:
101
102`system/workflow -> validation -> main result -> baseline comparison ->
103mechanism/diagnostic analysis -> application or generalization`
104
105Each subsection should have a claim-first opening and then data support.
106
107For ML/conference-style experiment sections, open `references/experiments.md`
108and make sure each major claim is backed by comparison, ablation, or stress-test
109evidence.
110
111### Related Work
112
113Use:
114
115`topic scope -> representative methods -> limitation tied to this paper ->
116distinction`
117
118Group prior work by technical topic and mechanism, not by publication year.
119
120### Discussion
121
122Use:
123
124`central advance -> evidence meaning -> relation to prior work -> constraints ->
125future use`
126
127This is where interpretation and limitations belong. Do not repeat the Results
128section figure by figure.
129
130### Conclusion
131
132Use:
133
134`contribution -> decisive evidence -> implication -> boundary`
135
136No new data. No unsupported promises.
137
138### Title
139
140Prefer concrete titles that combine:
141
142`system/object + action/capability + application or consequence`
143
144Avoid slogan titles, grant-style aims and overbroad field claims.
145
146## Output format
147
148Default output:
149
1501. `Draft:` with the requested prose.
1512. `Section outline:` with `3-7` compact bullets when the task involves a full section.
1523. `Assumptions or missing inputs:` with only material issues.
1534. `Claim-evidence map:` for major claims, using `Claim: ... | Evidence: ... | Status: supported/needs evidence`.
1545. `Why this structure:` with `2-4` short bullets.
155
156For Chinese author notes, provide polished English first, then brief Chinese
157notes explaining major structural choices.
Galaxy-Dawn