Quanxiang Liu
Community@Immortalqx · Hong Kong
Spatial Intelligence, Robotics
Agent Skills by Quanxiang Liu
Showing 28 vetted skills indexed across 1 GitHub repositories.
paper-reading
Produces structured critical reading notes from research paper PDFs with cropped figures and citations.
phd-benchmark-paper-template
Structures benchmark and evaluation papers using a five-pillar framework with section skeletons and checklists.
research-survey-loop
Runs multi-round literature surveys with task state files, chunked PDF reading, and incremental Chinese survey writing.
proof-writer
Writes rigorous mathematical proofs for ML and AI theory claims.
mock-review
Generates simulated venue-specific peer reviews of manuscripts with PDF artifact scans and literature grounding.
theme-factory
Applies curated color palettes and font pairings to slides, documents, and HTML artifacts.
formula-derivation
Structures research formula derivations from assumptions and problem statements into paper-ready theory drafts.
phd-vibe-research-workflow
Plans AI-assisted research sessions across coding, figure, and writing phases with integrity rules.
doc-coauthoring
Guides structured co-authoring of specs, proposals, PRDs, RFCs, and decision documents.
research-wiki
Builds and maintains a persistent per-project research knowledge base of papers, ideas, experiments, and claims.
xlsx
Create, edit, and recalculate Excel workbooks with openpyxl, pandas, and LibreOffice.
Read, merge, split, fill, and generate PDF documents using Python libraries and command-line tools.
phd-figure-designer
Designs and audits the three core figures of a technical research paper.
phd-pre-submission-reviewer
Reviews technical papers across five dimensions with severity-tagged findings before submission.
mmx-cli
Operate the local MiniMax CLI for web search, vision, and media generation tasks.
grant-proposal
Draft structured grant proposals for KAKENHI, NSF, NSFC, ERC, and other funding agencies.
phd-tech-paper-template
Builds a technical paper's logical skeleton with a thinking-template table and four consistency checks.
figure-description
Generate patent drawing descriptions and reference-numeral indexes from technical figures.
pptx
Create, edit, and inspect PowerPoint .pptx presentations with OOXML tooling and visual QA.
phd-intro-drafter
Generates a six-paragraph Introduction outline for technical research papers with aligned contributions.
arxiv
Search arXiv, fetch paper metadata, and download PDFs into a local library.
alphaxiv
Fetches structured AlphaXiv paper summaries with tiered fallback to arXiv LaTeX source.
drawio-diagram
Create editable draw.io research figures with PNG export and visual QA.
phd-idea-evaluator
Evaluates research ideas against a five-dimension scoring framework and returns a reviewer-style verdict.
Frequently Asked Questions About Quanxiang Liu
FAQPage SchemaWhat tasks can I accomplish with Quanxiang Liu's skill collection?▼
You can deeply read and summarize papers, run multi-round literature surveys, check research novelty, structure technical and benchmark papers, draft six-paragraph Introductions, write mathematical proofs, simulate venue-specific peer reviews, design paper figures, and produce PDF, DOCX, XLSX, and PPTX deliverables.
Who are these skills designed for?▼
They target PhD students, postdocs, and principal investigators in robotics, embodied AI, computer vision, world models, navigation, manipulation, and 3D scene understanding, plus anyone preparing conference submissions to venues like NeurIPS, CVPR, ICLR, ICCV, or ACM MM.
How do the skills handle research papers and citations?▼
The arxiv and alphaxiv skills search, download, and summarize papers; research-survey-loop migrates local PDFs into task-local sources, reads them in chunks of at most 10 pages, and maintains task.md, round_log.md, and survey.md files with relative local citations or web links.
Which grant funding agencies are supported?▼
The grant-proposal skill supports KAKENHI (Japan), NSF (US), NSFC (China, including 面上, 青年, 优青, 杰青, 海外优青, and 重点 programs), ERC (EU), DFG (Germany), SNSF (Switzerland), ARC (Australia), NWO (Netherlands), and generic formats.
What prerequisites or external services do these skills require?▼
Most skills operate on local files such as PDFs, .docx, .xlsx, and .pptx. The mmx-cli skill requires the local MiniMax CLI for web search, vision, and media generation; drawio-diagram exports PNG with a 60-second timeout; alphaxiv uses AlphaXiv summaries with tiered source fallback.