Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
By mronkko
Manage academic research workflows in Claude Code: enforce empirical integrity with pipeline-only data, manage Zotero citations and libraries, run PRISMA systematic reviews, and iterate manuscript revisions via parallel critics (evidence, method, argument, expert). Includes a terminal setup wizard.
npx claudepluginhub mronkko/claude-academic-research --plugin academic-researchUse when drafting or editing academic prose — topic sentences, paragraph structure, APA-style citations, active voice, tense, hedging conventions, term definitions. Fires eagerly on any .qmd / .Rmd / .md / .tex file in an academic-research project. Do NOT use for the revision workflow — use `manuscript-revision` + `/critic-loop`. Do NOT use for factual or citation accuracy — use `grounded-citations` / `empirical-integrity` / `fact-check`.
Use when the user asks to revise, polish, or critique an academic manuscript, chapter, thesis, or paper through parallel critic perspectives. Trigger phrases "/critic-loop", "run critic-loop", "revise this manuscript", "run critics on this draft", "run a critical review", "review this chapter". Invoke as `/critic-loop <document-path>`.
Use when editing manuscripts (.qmd, .Rmd, .ipynb, .tex, .md) that contain numbers, statistics, tables, test results, or methodology facts (search dates, model names, keyword strings). Enforces the rule that every quantitative or methodological claim in prose must come from a pipeline-generated authoritative file via code chunk or inline expression — never hand-typed.
Use when the user asks to fact-check a manuscript, verify citations, audit sources, or check whether cited papers actually support the claims attributed to them. Trigger phrases "fact-check", "verify citations", "audit citations", "check the sources", "do these papers actually say that", "verify the numbers in this draft". Do NOT use during or immediately after `/critic-loop` — the evidence critic inside that loop covers the same ground and burns MCP / Zotero quota twice.
Use when inserting a new citation into academic prose, attributing a finding to a source, or summarising what a paper says. Trigger phrases "cite this", "add a citation for X", "what does Smith (2019) say", "summarise this paper", "attribute this finding". Do NOT use for Zotero library housekeeping — use `zotero-operations`. Do NOT use for auditing an existing draft's citations or for verifying that a cited paper actually supports the claim — use `fact-check`.
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
9 research-hub skills: literature search, comparison matrix, planning manifests, design dialog, multi-AI routing, NotebookLM brief verification, paper-memory builder, Zotero curator. Auto-discovered from skills/<name>/SKILL.md.
Production-grade academic research pipeline for Claude Code: research → write → review → revise → finalize. 4 skills, 35+ modes, 38-agent ensemble, v3.7.3 + v3.8 L3 claim-faithfulness gate, v3.9.0 cross-index triangulation, v3.9.2 phase boundary fence (#133).
Semi-automated research assistant for academic research and software development, with skills for literature review, experiments, analysis, writing, and project knowledge management
Research integrity plugin for Claude Code — paper auditing, citation verification, experiment analysis, and methodology-first skills for academic workflows.
Research-team agents for Claude Code: supervisor, analysis-implementer, paper-writer, figure-descriptor, reviewer, literature-curator.
Citation, prose, figure, and rigor auditing for AI-assisted research drafts.
Claude Code plugin for academic research: MCP-grounded citations, empirical integrity, systematic reviews, Zotero operations, and parallel-critic manuscript revision.
Inside the Claude Code chat (Desktop or CLI):
/plugin marketplace add mronkko/claude-academic-research
/plugin install academic-research@mronkko
After install, run /setup once to configure API keys, MCP servers, and
permission rules. The wizard is chat-driven — no terminal required.
Works on Windows, macOS, and Linux. Windows users do not need WSL or
Git Bash — native cmd and PowerShell are supported, and CI verifies
every commit against windows-latest.
Eight skills:
| Skill | Mode | Purpose |
|---|---|---|
grounded-citations | rule-book (eager) | Every citation = Zotero BBT key + externalised source consultation; drop claims the source doesn't support. |
empirical-integrity | rule-book (eager) | Every number in prose must come from an inline expression reading analysis/results/. |
manuscript-revision | rule-book (eager) | Parallel-critic revision loop is the default revision protocol — delegates to /critic-loop. |
academic-style | rule-book (eager) | House-style conventions at drafting time — APA citations, voice, tense, hedging, synthesis-over-enumeration, terminology. |
systematic-review | procedure (explicit) | End-to-end SLR pipeline from search → screening → coding → export. |
zotero-operations | procedure (explicit) | Import, dedup, enrich, attach PDFs, maintain BBT keys. |
fact-check | procedure (explicit) | Verify citations and quantitative claims against sources. |
critic-loop | procedure (explicit) | Run 4 parallel critics (evidence / method / argument / expert) until no MAJOR issues remain. |
setup | procedure (explicit) | Chat-driven configuration wizard for first-time install. |
uv and use PEP 723 inline dependency declarations
— no venv, no requirements.txt, no pip install.~/.config/academic-research/config.toml (mode 0600) or
environment variables. Environment variables take precedence.permissions.deny rule blocks Claude's Read tool from accessing the
config file, so API keys never enter a conversation context.localhost:23119)..claude-plugin/
plugin.json # plugin manifest
marketplace.json # self-hosted marketplace catalog
skills/ # SKILL.md per skill
scripts/
core/ # llm, http, pdf, zotero primitives
sources/ # abstract/metadata sources (Crossref, Semantic Scholar, Scopus, ...)
publishers/ # per-publisher PDF retrieval (Wiley, Elsevier, ...)
pipelines/ # enrich_pdfs, enrich_abstracts, generate_bib, search, ...
setup/ # first-run configuration helpers
tests/unit/ # pytest + responses mocks
.github/workflows/ci.yml # pytest + ruff on push/PR
MIT. See LICENSE.