watericetangcw/academic-research-graph

academic-research-graph is a research Skill that also ships as a DeepSeek Harness profile bundle: give it a paper, concept, benchmark, claim or GitHub repository and it returns an interactive graph of the surrounding literature - lineage, per-paper guides, criticism and contradiction edges, verified benchmark adoption - that you can extend node by node.

academic-research-graph researches the literature around one input - a paper URL, DOI, arXiv ID, title, PDF, a concept such as Vision Transformer, a benchmark or dataset, a concrete claim, or a paper-associated GitHub repository - and returns an interactive graph rather than a citation list. Nodes are clickable papers carrying preprint / conference / journal status and sourced venue badges; edges trace the main intellectual lineage, and separate conflict edges mark criticism, contradictory findings, failed reproductions and cases where two papers only appear to disagree because of different data or evaluation protocols. Each important node carries a readable guide (background, problem, method, evidence, significance, limitations) for someone who knows the field but has not opened the PDF. Benchmark adoption is verified rather than assumed: the graph distinguishes a paper that merely cites a benchmark from one that evaluates on it, loads its data or uses its official code. From a repository it resolves the associated paper, inspects the implementation at a pinned commit and compares code with the described method; any existing node can seed a new branch whose reviewed findings merge back into the same graph.

Learning & Research ★ 5
View on GitHub ↗

Install

npx @deepseek-ai/dsh plugin --profile web add github:watericetangcw/academic-research-graph

Copied verbatim from the README's DeepSeek Harness section (the README prefixes the command with npx @deepseek-ai/dsh after corepack enable). No npm package is documented for the bundle, so the GitHub form is shipped.

Compatibility

The README lists Node.js ^22.19.0 or >=24.0.0 and pnpm on PATH (corepack enable can provide it) as prerequisites, then npx @deepseek-ai/dsh web to start Harness. The same repository also installs as a ChatGPT Skill, and the graph output can be produced in Chinese, English or bilingual mode via output_language. Generated graphs are static pages committed to the examples folder, with a live example hosted on GitHub Pages.

Details

Alternatives

1692775560/dsh-Mimir-Academic-research · Astervolans/dsh-literature-search · beibingyangliuying/dsh-autoresearch

More plugins in Learning & Research

Browse more in Learning & Research

Guides for Learning & Research plugins