imbad0202/academic-research-skills
4 skills11.9k installs115k starsGitHub
Install
npx skills add https://github.com/imbad0202/academic-research-skillsSkills in this repo
1Academic Paper ReviewerAcademic Paper Reviewer is an agent skill that replays an international journal peer review in one session. It identifies the manuscript’s field, assembles five reviewers—including a Devil’s Advocate who stress-tests claims and logic—and returns referee-style feedback plus an editorial decision and revision plan. Solo builders and small research teams use it when they want harsh, structured critique before arXiv upload, conference submission, or journal resubmission, without waiting on human reviewers. Modes span a full committee review, targeted methodology passes, quick triage, guided Socratic questioning, re-review to verify fixes, and calibration to compare simulated vs. real reviewer tone. It pairs naturally with academic writing and pipeline skills in the same repo for drafting through submission prep. Install it when you need consistent, multi-angle manuscript QA rather than a single generic “proofread this” pass.3.7kinstalls2Academic PaperThe academic-paper skill fragment documented here is the Abstract Bilingual Agent inside imbad0202’s academic-research-skills workflow. Solo builders and researchers using agentic paper pipelines invoke it at Phase 5b to finalize bilingual abstracts: one English and one Traditional Chinese version, each written as native-quality prose rather than a literal translation of the other, with matched keyword sets. It assumes prior phases already produced configuration, literature, structure, arguments, and a full draft; it may read those directories but must only write Phase 5 bilingual abstract deliverables. It deliberately stops before peer review (Phase 6) and formatting (Phase 7), handing control back to the orchestrator if quality issues need downstream agents. Use when you are automating thesis or journal submission prep with strict phase gates—not for general blog posts or English-only summaries.3kinstalls3Deep Researchdeep-research packages the Bibliography Agent for academic and evidence-heavy solo work: you get a disciplined Phase 2 (Investigation) pass that finds sources, applies criteria, annotates them in APA 7.0, and records how the search was run so someone else can repeat it. The skill is intentionally narrow—unlike open-ended “deep research” chat, it refuses to inflate into synthesis chapters, drafting, peer review, or revision folders, which protects multi-phase pipelines from cross-contamination. Install it when you already have or are building phase1 research briefs and need citable literature rows before validation or build. It fits agentic dissertation or grant workflows where each phase owns one artifact type. Pair it with later-phase skills only after the bibliography deliverable is accepted, not in the same invocation.2.7kinstalls4Academic PipelineClaim Reference Alignment Audit Agent is the L3 auditor in the Academic Research Skills (ARS) pipeline. Solo builders and small research teams use it when a Stage 4 draft already contains citations and they must judge each claim against the retrieved text of the cited reference—not trust locator markers alone. The agent produces evidence-bound verdicts, buckets uncited and drifting assertions, and feeds passport aggregates so the Stage 5 formatter can hard-stop on substantive faithfulness failures. It does not rewrite the paper or override formatter policy; it documents defects with severity tiers aligned to recent literature on hallucinated citations. Placement is multi-phase: primary shelf Ship/review for pre-publication QA, with Build/docs as the drafting context where claims are introduced. Expect advanced pipeline literacy (stages, passports, external spec URLs). Pair with upstream retrieval/locator skills and downstream formatting skills in the same repo stack.2.5kinstalls