What this resource does
Core uses
This repository is a collection of separate Skills
The reviewed repository lists 19 user-facing Skills and a separate shared support Skill. This page presents 17 selected Skills grouped by task, with their documented names.
The selection omits nature-paper-card and nature-image2ppt; shared support files are not standalone task choices.
Paper reading and research notes
Turns a supplied paper into a structured, figure-aware bilingual Markdown reader with source anchors.
- Typical inputs
- Paper PDF, DOI, arXiv link, publisher page, or pasted manuscript text; Preferred reading or translation focus
- Output
- A bilingual Markdown reading document with aligned figures and source references.
Best for: Close reading, literature notes, and journal-club preparation.
Skill for this task
nature-reader
Use it to read a paper and create structured bilingual notes.
Suggested requests (3)
Turn this PDF into a full Chinese-English Markdown reader.
Translate and explain this paper, placing figures near the relevant text.
Use this DOI to get the paper content and generate a reader with a source map.
From skills/nature-reader/README_EN.md.
Academic writing and polishing
Drafts manuscript sections from research materials or revises supplied academic prose for structure, language, and journal style.
- Typical inputs
- Claims, results, figures, notes, or an existing draft; Target section, journal, and terminology constraints
- Output
- Draft or revised manuscript text, with the exact deliverable depending on the selected component.
Best for: Building manuscript sections or improving an existing academic draft.
Skills for this task
nature-writing
Use it to draft manuscript sections from research materials.
Suggested requests (3)
Write a Nature-style abstract from these figures and results.
Rebuild the logic of this introduction; do not only polish the sentences.
Turn these Chinese results into an English Results narrative.
From skills/nature-writing/README_EN.md.
nature-polishing
Use it to revise academic prose for structure, language, and journal style.
Suggested requests (3)
Translate this Chinese Results paragraph into Nature-style English, keeping it conservative.
Polish this abstract without changing facts or citation intent.
This introduction sounds too AI-written; rebuild its logic and language.
From skills/nature-polishing/README_EN.md.
Peer review and revision response
Assesses manuscripts from reviewer perspectives or prepares point-by-point revision correspondence from reviewer comments.
- Typical inputs
- Manuscript or revision package; Reviewer comments and author response evidence
- Output
- Reviewer reports, response letters, revision cover letters, or marked revision plans.
Best for: Pre-submission review and responding to journal revision decisions.
Skills for this task
nature-reviewer
Use it to examine a manuscript from a reviewer's perspective.
Suggested requests (3)
Review this introduction and Figure 1 like a Nature reviewer.
Before submission, find the technical issues reviewers are most likely to attack.
Give me three reviewer reports and one synthesis, not a rebuttal.
From skills/nature-reviewer/README_EN.md.
nature-response
Use it to prepare point-by-point replies to reviewer comments.
Suggested requests (3)
Here is the editor email and reviewer comments; generate a point-by-point response framework.
Turn my Chinese revision notes into English reviewer responses.
Check whether this rebuttal misses anything, sounds too strong, or lacks evidence.
From skills/nature-response/README_EN.md.
Literature and reference work
Searches academic sources, audits references, and exports selected citation records in documented bibliographic formats.
- Typical inputs
- Research topic, DOI, citation, or reference list; Selected database or citation scope
- Output
- Search results, citation tables, reference-check findings, and ENW, RIS, or Zotero RDF records.
Best for: Literature discovery, reference checking, and citation-library preparation.
Skills for this task
nature-academic-search
Use it to search academic sources for literature relevant to a topic.
Suggested requests (4)
Check the strict external citations for this paper and list influential citers.
Find 20 recent Nature/Science/Cell-related papers on this topic.
Convert these DOIs into Nature-style references and export RIS.
Add MeSH terms and synonyms to this PubMed query.
From skills/nature-academic-search/README_EN.md.
nature-citation
Use it to save selected references in documented citation formats.
Suggested requests (3)
Split this introduction paragraph and add Nature-series citations.
Use only CNS and subjournal papers from the last five years to support these claims.
I have confirmed these DOIs; export them in a Zotero-importable format.
From skills/nature-citation/README_EN.md.
nature-ref-verifier
Use it to compare and check bibliographic details.
Suggested requests (3)
Check these 50 references one by one and mark the ones that must be fixed.
These DOIs may be wrong; find the real titles and page ranges.
Turn the incorrect fields in this BibTeX file into patch suggestions.
From skills/nature-ref-verifier/README_EN.md.
nature-literature-pipeline
Use it for a repeatable or scheduled literature-discovery workflow.
Suggested requests (3)
Track new papers on marine concrete chloride diffusion and machine learning every morning.
Create a weekly literature pipeline for this keyword set.
Score papers from the last seven days, keep the Top 5, and archive them.
From skills/nature-literature-pipeline/README_EN.md.
Figures and presentations
Builds scientific figures or turns a research paper into a Chinese paper-presentation deck.
- Typical inputs
- Paper, data, method description, or figure requirements; Presentation purpose or visual constraints
- Output
- Scientific figure assets or a PPTX presentation with figures, notes, and a quality report.
Best for: Manuscript visuals, group meetings, journal clubs, and paper presentations.
Skills for this task
nature-figure
Use it to create scientific figure assets from data or method requirements.
Suggested requests (4)
Make a Nature-style multi-panel figure from this dataset, preferably in Python.
Use the figures4papers Nature Machine Intelligence layout as a reference and add a method-comparison figure.
Redraw this mechanism schematic, export SVG/PDF, and give me the source-data table.
Use OpenRouter to draft a graphical abstract, but do not treat it as a quantitative data figure.
From skills/nature-figure/README_EN.md.
nature-paper2ppt
Use it to turn a research paper into a Chinese PPTX presentation.
Suggested requests (3)
Turn this Nature Communications paper into a 12-slide Chinese group-meeting PPT.
Use the abstract and figure legends to make a first paper-presentation draft; keep it concise.
Make this machine-learning paper into a deck with clear method, results, and limitations.
From skills/nature-paper2ppt/README_EN.md.
Data statements and statistics
Prepares data-availability materials or audits statistical reporting for research manuscripts.
- Typical inputs
- Data-sharing context, repository details, or an existing availability statement; Statistical methods, results, figure legends, or supplied data when computation is requested
- Output
- A Data Availability statement, repository or FAIR checklist, statistical reporting audit, or revised statistics section.
Best for: Checking data-sharing and statistical-reporting sections before manuscript submission.
Skills for this task
nature-data
Use it to prepare data-availability statements, repository plans, and FAIR metadata checks.
Suggested requests (3)
Write this manuscript's Data Availability statement in Nature style.
Some data cannot be public; draft a controlled-access statement.
Check whether my data statement is missing accessions, repositories, or licenses.
From skills/nature-data/README_EN.md.
nature-statistics
Use it to audit or revise statistical reporting in a research manuscript.
Suggested requests (3)
Check whether this Statistical analysis paragraph is Nature-style.
How should I write the error bars, n, and p values in this figure legend?
The reviewer says the statistics are insufficient; list revision options and draft a response.
From skills/nature-statistics/README_EN.md.
Authorised paper access
Uses legitimate open-access routes or the user's authorised library session to find and download academic full text.
- Typical inputs
- Paper title, DOI, URL, or search details; An authenticated institutional browser session when subscription access is required
- Output
- An authorised full-text file when available, together with a clear access or download result.
Best for: Obtaining papers through open-access or institution-authorised routes.
Skill for this task
nature-downloader
Use it to obtain authorised or open-access academic full text through documented access routes.
Suggested requests (3)
Configure my university library entry point so future paper downloads can reuse it.
Use my logged-in Chrome session to download PDFs for these DOIs into the current project.
Use the authorized CNKI route for this Chinese paper; save it if possible and explain why if not.
From skills/nature-downloader/README_EN.md.
Experiment records
Converts experiment images, voice transcripts, and text notes into standardised research logs in an Obsidian vault.
- Typical inputs
- Experiment images, voice transcripts, or text notes; Obsidian vault and optional Feishu workflow details
- Output
- A structured experiment log with YAML frontmatter and archived source materials.
Best for: Turning raw experiment records into consistent, searchable lab notes.
Skill for this task
nature-experiment-log
Use it to turn experiment images, voice notes, and text into structured Obsidian logs.
Suggested requests (3)
Record an experiment: 316L chloride-salt corrosion at 500°C for 300 h in Ar, mass loss 0.0032 g.
Turn these experiment photos into today's Obsidian lab log.
Use this voice transcript to create a standard experiment record and list missing information.
From skills/nature-experiment-log/README_EN.md.
Research proposals and patents
Structures evidence and arguments for research proposals or converts technical research materials into Chinese invention-patent drafts.
- Typical inputs
- Research evidence, project notes, objectives, or an existing proposal draft; Paper, thesis, technical report, source code, figures, or manuscript for patent work
- Output
- A research proposal draft and QA structure, or an evidence-grounded Chinese patent drafting package.
Best for: Developing research proposals or documenting supported technical contributions for patent drafting.
Skills for this task
nature-proposal-writer
Use it to plan and draft research proposals from an evidence and argument structure.
Suggested requests (3)
I have a grant title; first break down the scientific questions and research content, do not write prose yet.
This introduction is scattered; rebuild it with the proposal-first workflow.
Expand this draft into a proposal framework and mark evidence gaps.
From skills/nature-proposal-writer/README_EN.md.
nature-paper-to-patent
Use it to convert research materials into evidence-grounded Chinese invention-patent drafts.
Suggested requests (5)
Convert this paper into a Chinese invention patent disclosure and claim draft.
Scan this project, identify patent points, and draft a technical disclosure.
Merge these new materials into the existing disclosure and save a new Word version.
Extract only the patentable points first; do not write the full specification yet.
Compare the paper claims and patent claims, and find features without evidence.
From skills/nature-paper-to-patent/README_EN.md.
Use boundaries
Limits and checks
Verify evidence before use
Generated citations, claims, calculations, or reviewer comments may contain mismatched details.
Compare important statements, DOIs, quotations, and numbers with the original source.
Check the selected Skill's status
The official repository labels components as Stable, Beta, or Draft, so behaviour may differ between Skills.
Try Beta or Draft components on non-critical material before relying on their output.
Review access and local requirements
Some Skills use APIs, accounts, browser sessions, local scripts, or external services.
Check costs, data policies, credentials, dependencies, and file changes for the selected Skill.
More boundaries
- Task boundary: It is not designed to complete every research and publication task through one selected module.
- Input boundary: It cannot analyse inaccessible subscription content unless the user has legitimate access and supplies the required material.
Technical details
- Resource type
- Skill Collection
- Author or maintainer
- Yizhe Yuan
- Source last updated
- 13 Jul 2026
- Last verified
- 13 Jul 2026
- Licence
- Apache-2.0
- Access
- Publicly available
- Additional costs
- Platform terms or usage limits may apply. API usage fees may apply for selected components. External service or library-access terms may apply for selected components.
- Skill instruction language
- English, Chinese
- Documentation language
- English, Chinese
- Repository languages
- Markdown, Python, Shell
- Dependencies
- A documented supported agent platform; Git for the recommended clone-based setup; Module-specific Python packages, APIs, credentials, shared files, or browser access
- Review status
- Quick test completed
Continue exploring
Related resources
Official package contents
Files
Read the official English instructions without leaving this profile. This view includes the English files represented by this profile, plus the licence, English agent setup notes, and Codex installation script. Non-English source files, supporting code, assets, tests, and references may not be shown.
Official package files
20 files available here
skills17
nature-academic-search1
nature-citation1
nature-data1
nature-downloader1
nature-experiment-log1
nature-figure1
nature-literature-pipeline1
nature-paper-to-patent1
nature-paper2ppt1
nature-polishing1
nature-proposal-writer1
nature-reader1
nature-ref-verifier1
nature-response1
nature-reviewer1
nature-statistics1
nature-writing1
docs1
scripts1
Read on this page
Select a file
Choose a file from the tree to read its official instructions without leaving this page.
Independent testing
Evaluation report
See exactly what was tested, which checks were used, and what the result does not cover.
Evaluations are mainly based on automated AI-assisted checks. Results are for reference only.
Component totals mentioned in these saved examples refer to the inventory used for those tests, not the current Profile inventory. The results cover only the named components.
Evaluation coverage
4 components tested separately
Each score applies only to the named component, cases, application and model. The scores are not combined into a Repository rating.
Revised examples have a separate final-presentation score, not a first-attempt execution score.
13 named components were not tested
nature-writingnature-polishingnature-reviewernature-responsenature-citationnature-ref-verifiernature-literature-pipelinenature-datanature-statisticsnature-downloadernature-experiment-lognature-paper-to-patentnature-proposal-writer
Tested component 1nature-academic-search95 / 100
We supplied a correct reference and a copy with two deliberate mistakes. The component checked Crossref and PubMed and corrected the year from 2020 to 2019 and the article number from 5234 to 5233. It also found the paper's PubMed identifier, but confusingly called that identifier unresolved in its summary.
This score covers one example only. It does not rate every function or predict results on other materials.
Capability types: Citation And Reference
Example results
What was provided and what came back
Example 1
Check two references for bibliographic errors
What we asked it to do
Check the two supplied references using the official citation-verification route. Compare title, ordered authors, journal, year, volume, article number, DOI and PMID. Return a concise English table with the supplied value, source value and any correction.
Original material
Excerpt from final result
What this test shows
- Reference 1 matches the saved records: 2019, volume 9, article 5233, DOI 10.1038/s41598-019-41695-z and PMID 30914743.
- Reference 2 uses the same title, authors, journal, volume and DOI but incorrectly supplies year 2020 and article number 5234; the saved records give 2019 and 5233.
- The saved PubMed record associates PMID 30914743 with the same DOI and paper.
- The output’s Reference 2 PMID row displays 30914743 but calls it unverified because it was absent from the supplied version.
This example demonstrates that the component can detect the two deliberately incorrect citation fields using saved public records, while also showing a reporting error around an identifier that was absent from the input but present in PubMed.
Score breakdown
Five checks for this component
Did it complete the task?
100%The planned task and output checks recorded 4 passed, 0 partly passed, and 0 failed.
Did it run without problems?
75%The setup, dependency, execution, and runtime checks recorded 2 passed, 2 partly passed, and 0 failed.
Did it preserve and handle the supplied information responsibly?
100%The checks on supplied information, evidence, and academic boundaries recorded 4 passed, 0 partly passed, and 0 failed.
Did it do the documented task in the tested setup?
100%The checks for the documented function in the recorded setup recorded 4 passed, 0 partly passed, and 0 failed.
Were the setup and use instructions clear enough to follow?
100%The setup and use-instruction checks recorded 4 passed, 0 partly passed, and 0 failed.
Problems found
What happened in these examples
- Reference 2’s PMID row displays the retrieved PMID but says the source did not provide it; the source did provide it, while the supplied reference did not.
- This run contacted Crossref and PubMed directly. The optional agent-to-database connection described in the documentation was not installed or tested.
What was not tested
- One case only. Results and any score apply to this example, not to all uses of the component.
- Topic searches, MeSH construction, citation-network analysis, bibliography conversion and database coverage beyond the sources actually reached.
- Only two supplied versions of one real reference were tested. Topic searches, MeSH construction, citation-network analysis, bibliography conversion, arXiv, Semantic Scholar, Scopus, ScienceDirect, OpenAlex discovery, configured APIs and account-based services were not tested.
- This review covers four of the repository's 17 components, with one example per component. The other 13 components were not tested in this review.
- Repository components not tested in this review: nature-writing, nature-polishing, nature-reviewer, nature-response, nature-citation, nature-ref-verifier, nature-literature-pipeline, nature-data, nature-statistics, nature-downloader, nature-experiment-log, nature-paper-to-patent, nature-proposal-writer.
Tested component 2nature-figure96.3 / 100
A three-panel figure was created from the supplied simulated measurements in PNG, SVG and PDF. The plotted values, confidence limits, day-14 subsets, colors, units, shared legend and missing-value disclosure matched the input. Visual inspection found no clipping, overlapping labels or missing uncertainty bands; the only requested-detail difference was lowercase a/b/c panel labels instead of uppercase A/B/C.
This score covers one example only. It does not rate every function or predict results on other materials.
Capability types: Data Visualisation
Example results
What was provided and what came back
Example 1
Plot a three-panel dose-response figure
What we asked it to do
Use Python to create a 180 mm by 80 mm, three-panel manuscript figure from the supplied simulated measurements. A: time courses with shaded confidence intervals; B: day-14 estimates with error bars; C: dose versus day-14 response. Use one shared legend, panel labels A/B/C, explicit units and a simulated-data caption.
Original material
Final result
What this test shows
- The PNG is 4248 by 1884 pixels and visibly contains the time-course, day-14 group and dose-response panels.
- Panel A contains nine measured points and three supplied confidence bands; no day-21 point is drawn.
- Panels B and C each contain the three estimates 1.2, 1.8 and 2.6 with intervals 1.0–1.4, 1.4–2.2 and 2.0–3.2.
- The caption identifies simulated data and discloses the excluded High dose day-21 measurement.
This example demonstrates that the component can turn the bounded dataset into the requested multi-panel figure without changing or inventing measurements, with a minor mismatch in panel-label capitalization.
Score breakdown
Five checks for this component
Did it complete the task?
87.5%The planned task and output checks recorded 3 passed, 1 partly passed, and 0 failed.
Did it run without problems?
100%The setup, dependency, execution, and runtime checks recorded 4 passed, 0 partly passed, and 0 failed.
Did it preserve and handle the supplied information responsibly?
100%The checks on supplied information, evidence, and academic boundaries recorded 4 passed, 0 partly passed, and 0 failed.
Did it do the documented task in the tested setup?
100%The checks for the documented function in the recorded setup recorded 4 passed, 0 partly passed, and 0 failed.
Were the setup and use instructions clear enough to follow?
100%The setup and use-instruction checks recorded 4 passed, 0 partly passed, and 0 failed.
Problems found
What happened in these examples
- The visible panel identifiers are lowercase a/b/c, while the locked request specifies A/B/C.
What was not tested
- One case only. Results and any score apply to this example, not to all uses of the component.
- R plotting, image-generation APIs, interactive charts, other chart families and other datasets.
- Only one simulated Python figure with three specific panel types was tested. R plotting, image-generation APIs, interactive charts, fitted or smoothed models, TIFF export, other chart families and other datasets were not tested.
- This review covers four of the repository's 17 components, with one example per component. The other 13 components were not tested in this review.
- Repository components not tested in this review: nature-writing, nature-polishing, nature-reviewer, nature-response, nature-citation, nature-ref-verifier, nature-literature-pipeline, nature-data, nature-statistics, nature-downloader, nature-experiment-log, nature-paper-to-patent, nature-proposal-writer.
Tested component 3nature-reader100 / 100
We supplied five short sections of a fictional research paper. The component returned the original English beside a Chinese translation, preserving all five sections, the numbers and the equation. It kept the warning that one donor cohort cannot establish cause and effect, and did not invent missing publication or instrument details.
This score covers one example only. It does not rate every function or predict results on other materials.
Capability types: Information Extraction
Example results
What was provided and what came back
Example 1
Create a bilingual reader without changing the reported results
What we asked it to do
Use the official pasted-text route to create a complete English-Chinese Markdown reader of the supplied short research text, not a summary. Preserve source blocks, the equation and every numeric value. Produce paper.md, source_map.json and translation_notes.md and run the documented math validator.
Original material
Final result
What this test shows
- All five sections appear with the original English beside a Chinese translation.
- The equation is preserved as editable text rather than replaced with an image.
- The translation keeps 24 samples, 14 days, a 1.8-fold level and the distinction between that level and an 80 percent increase.
- The component's formula-format check reported no problems.
This example demonstrates that the component can produce a complete bilingual reader from a short pasted research text while preserving its quantities, equation and cautious conclusions instead of reducing it to a summary.
Score breakdown
Five checks for this component
Did it complete the task?
100%The planned task and output checks recorded 4 passed, 0 partly passed, and 0 failed.
Did it run without problems?
100%The setup, dependency, execution, and runtime checks recorded 4 passed, 0 partly passed, and 0 failed.
Did it preserve and handle the supplied information responsibly?
100%The checks on supplied information, evidence, and academic boundaries recorded 4 passed, 0 partly passed, and 0 failed.
Did it do the documented task in the tested setup?
100%The checks for the documented function in the recorded setup recorded 4 passed, 0 partly passed, and 0 failed.
Were the setup and use instructions clear enough to follow?
100%The setup and use-instruction checks recorded 4 passed, 0 partly passed, and 0 failed.
What was not tested
- One case only. Results and any score apply to this example, not to all uses of the component.
- PDF extraction, OCR, publisher retrieval, figure extraction and follow-up questions. No PDF export was requested or tested.
- Only one short evaluator-created pasted-text sample with five blocks and one equation was tested. PDF extraction, OCR, publisher retrieval, DOI or arXiv acquisition, figures, tables, raster page images, PDF export and follow-up questions were not tested.
- No real published paper, external citation, publisher account, retrieval service or user-provided private material was used.
- This review covers four of the repository's 17 components, with one example per component. The other 13 components were not tested in this review.
- Repository components not tested in this review: nature-writing, nature-polishing, nature-reviewer, nature-response, nature-citation, nature-ref-verifier, nature-literature-pipeline, nature-data, nature-statistics, nature-downloader, nature-experiment-log, nature-paper-to-patent, nature-proposal-writer.
Tested component 4nature-paper2ppt96 / 100 · Revised
Revised after feedback
Final presentation score
96 / 100This score reviews the finished presentation after feedback. It is not the original Skill run score, a visual-style rating, or a rating of the whole collection.
The presentation meets the ten-slide English brief, and its headline result and chart values match the paper. Four points were deducted because enzyme-activity and stoichiometry results are not explained. Runtime reliability and first-attempt performance were not scored.
How this score was calculated
Content and numerical accuracy 90%
Weight: 40%
The headline result and all 12 chart values match the paper. Enzyme-activity and stoichiometry findings are not explained in the slides, so coverage is partial.
Research presentation structure 100%
Weight: 25%
The ten slides move from the experiment to results, interpretation and two specific discussion questions.
File and presentation usability 100%
Weight: 20%
The PPTX opens in the inspection library, has English slide text and notes on all ten slides, and contains three editable charts. PowerPoint itself was not tested.
Sources and evidence limits 100%
Weight: 15%
Source labels are present, concept drawings are labelled, and conflicting values in the paper are disclosed rather than silently resolved.
We used the same 14-page paper to make a ten-slide English group-meeting presentation. After the first deck looked too plain, we prepared three design samples. The user approved that direction, then we completed the ten-slide version shown below.
This is a revised example made with user feedback in Codex, with additional AI-generated concept illustrations and a new layout. The score below applies only to this finished presentation, not the first attempt or the earlier execution test.
A ten-slide presentation from a wastewater-treatment paper
Original material
User-supplied 14-page research paper. Enhancing removal performance of ciprofloxacin hydrochloride by Chlorella vulgaris-bacteria consortium: Insights of microbial responses and stoichiometry. Yongrui Pi, Yan Li, Wenpeng Jia and Yongzheng Tang. Journal of Hazardous Materials 494 (2025), 138780. DOI: 10.1016/j.jhazmat.2025.138780. Task: turn the supplied PDF into a ten-slide English research presentation with source figures and speaker notes.
What we asked it to do
Turn this paper into a ten-slide English presentation for a research group meeting. You do not have to retain the original figures; you may crop them or redraw explanatory diagrams.
Final presentation
10 English slides, with speaker notes.
Download revised PowerPoint
What changed
- The final file contains ten English slides and English speaker notes. The slides cover the experiment, removal results, kinetics, physiological responses, microbial communities, a proposed mechanism, limitations and discussion.
- The result slide highlights 94.6 ± 0.6% removal at 10 mg/L after seven days and compares the other tested concentrations. Three charts are editable in PowerPoint.
- Growth, pigment and community figures now use Figures 2, 3 and 5 with matching source labels. The incorrect figure references from the earlier deck are not carried over.
- The cover and mechanism slides use labelled AI-generated concept illustrations. The other results retain the paper's charts, axes and legends.
- The user approved the three-slide design sample and then the complete ten-slide deck. That approval is a design preference, not a new performance score.
The main result is easier to find because it has its own large number and concentration chart. The mechanism uses labelled cells and arrows instead of a text-only explanation. Dense research figures remain in their original form so their labels and data can still be read. These improvements came from the revision process, not from changing the paper's findings.
Read the slide text
What this example does not show
- One paper and one ten-slide English presentation. This example does not establish results for other papers, languages or presentation lengths.
- The finished design includes user feedback and additional image generation. It does not show what the original short request produces without follow-up.
- The cover and cell drawings are conceptual illustrations, not experimental images. The mechanism arrows describe proposed exchanges, not measured exchange rates.
- The paper contains conflicting control values and sampling descriptions. The deck flags these rather than choosing a supposedly correct version.
- The final-presentation score is a retrospective content and file review. It does not grade visual taste, first-attempt execution, installation, or runtime reliability, and is not comparable to the other component execution scores.
Want to see more tests?
Tell us which module or task you would like us to test next.