scholaraio

Nature Workflow Quick Start

/nature-workflow is ScholarAIO’s bridge to the upstream nature-skills bundle. Use it when the task is explicitly Nature Skills, Nature-style, Nature-family/CNS, high-impact journal, or Springer Nature oriented.

It is not a simplified local clone. When the original upstream nature-* skill is installed or loadable, use that skill directly. If it is unavailable, ScholarAIO routes to local fallback skills and states that the output is a fallback, not an upstream-equivalent Nature Skills result.

When To Use It

Use /nature-workflow for:

• Nature/high-impact scientific figures and publication plots
• Nature-style manuscript polishing or Chinese-to-English academic polish
• Nature-style manuscript section drafting and argument restructuring
• Reviewer-style critique before submission
• Nature/CNS citation support and reference-manager export
• Nature/high-impact Data Availability, repository, source data, code availability, and FAIR checks
• Full-paper bilingual Markdown readers and source-grounded translation
• Reviewer response or rebuttal letters
• Paper-to-PPT / Chinese journal-club decks
• Nature-focused academic search, citation verification, MeSH strategy, and reference management

Submission package is only one scenario. Do not require a submission package when the user only asks for a figure, polish, reader, paper2ppt deck, citation search, or Nature Data Availability wording.

Inputs To Provide

Give the bridge the concrete artifact for the chosen route:

• Figure work: data table, result claim, target backend (Python or R), and export needs
• Polishing or writing: draft text, section type, target journal/style, and claims that must not change
• Reader or PPT: PDF, DOI, arXiv URL, publisher URL, or existing paper notes
• Citation/search: claim text, scope, preferred citation format, and any journal constraints
• Data Availability: dataset list, access restrictions, repositories, code status, and accession IDs if available
• Response/reviewer work: reviewer comments, decision letter, manuscript draft, and revision notes

Install Upstream Skills

For direct upstream behavior in Codex, install the full upstream bundle:

codex plugin marketplace add https://github.com/Yuan1z0825/nature-skills --ref main
codex plugin add nature-skills@nature-skills

For manual local-skill installation, clone the upstream repository and copy whole directories:

git clone https://github.com/Yuan1z0825/nature-skills.git
cd nature-skills
mkdir -p ~/.codex/skills
cp -R skills/_shared ~/.codex/skills/
for d in skills/nature-*; do
  cp -R "$d" ~/.codex/skills/
done

Copying only SKILL.md is not enough.

Common Scenarios

Scenario	Upstream route	ScholarAIO fallback if upstream is unavailable
Nature-style figure	nature-figure	/draw
Manuscript polish	nature-polishing	/writing-polish
Section drafting	nature-writing	/paper-writing
Mock reviewer critique	nature-reviewer	bounded fallback critique, then /paper-writing or /citation-check for fixes
Citation support	nature-citation	/search + /citation-check + /export
Data Availability	nature-data	gap inventory, then /paper-writing or /writing-polish wording
Full-paper reader	nature-reader	/show + /translate
Reviewer response	nature-response	/review-response
Paper-to-PPT	nature-paper2ppt	/show or /paper-guided-reading, then /document
Academic search	nature-academic-search	/search, /websearch, /citation-check, /export

Example Prompts

/nature-workflow Polish this abstract in Nature Communications style.

/nature-workflow Use Python to create a Nature-style multi-panel figure from these benchmark results.

/nature-workflow Turn this PDF into a Chinese journal-club PPT.

/nature-workflow Draft a Data Availability statement and repository plan for this manuscript.

/nature-workflow Simulate a Nature-style reviewer critique for this draft.

Expected Outputs

The first response should identify:

1. Upstream target: selected nature-* skill or skill sequence.
2. Mode: direct upstream use or ScholarAIO fallback.
3. Route: exact route and fallback skills if needed.
4. Immediate next action: file to inspect, command to run, or one blocking question.
5. Guardrails: source grounding, no invented citations, no invented data identifiers, and required verification.

Production Demo

The tracked demo fixture lives at tests/fixtures/nature_workflow_demo/. Run its verifier with:

python tests/fixtures/nature_workflow_demo/verify_demo.py

The fixture also includes an executable route demo. It classifies practical prompts, writes route cards, and proves both direct-upstream and fallback modes:

python tests/fixtures/nature_workflow_demo/run_demo.py --output route-cards.md

For a research-output-level demo, generate the full artifact package:

python tests/fixtures/nature_workflow_demo/run_product_demo.py --output-dir workspace/_system/issue-107-routing-eval/nature-workflow-product-demo

That package includes route-cards.md, inputs/source-data.csv, SVG/PDF/PNG figure exports, a polished abstract, a Data Availability statement, a generated PPTX deck, and QA reports.

The verifier checks that the bridge covers the full upstream 10-skill index, records the direct-use policy, executes the route demo, and does not collapse the workflow into a submission-only package.

Guardrails

• Use the original upstream nature-* skill directly when available.
• Do not emulate upstream behavior from a short local summary.
• Do not copy only SKILL.md; use the whole skill directory and skills/_shared when needed.
• Mark ScholarAIO routes as fallbacks when upstream is unavailable.
• Do not invent citations, reviewer expectations, journal rules, data IDs, repository URLs, source data, figures, or manuscript claims.

This site is open source. Improve this page.