Open Scholar Skill plugin
35 skills for Claude Code and Codex lead a social scientist from idea to a journal manuscript, with scans of participant data.
- 190 ★ on the repository
- Last commit
Broad access
Why this level
- The setup script changes agent configuration and registers hooks that intercept file reads and tool output.
- Data read through the agent is sent to the model API, and protected datasets need local mode.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/open-scholar-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The setup script changes agent configuration and registers hooks that intercept file reads and tool output.
- Data read through the agent is sent to the model API, and protected datasets need local mode.
This is third-party code. Review the repository files before installing.
What it does
Open Scholar Skill covers the whole research path in the social sciences: idea, literature review, hypotheses, design, causal inference, data work, analysis, writing, citations, verification, journal choice, replication and simulated peer review. It is 35 skills and 20 agents (nine reviewers, five for verification and six for code review) plus a `sync-docs` utility. Skills run one at a time, and the author advises staying in the loop at each step instead of handing research to autopilot. A separate data-protection layer works through `/scholar-init`: a local scan for personal data, access statuses and hooks that intercept reads of protected data files. Installation is a `setup.sh` script that detects Claude Code, Codex and ZCode and puts skills into the right folders.
Who it is for. Sociologists and other social scientists preparing articles for top journals who work with survey or interview data.
Good fit when
- A literature review must lead to testable hypotheses.
- Before working with IRB-protected interviews, data is scanned for personal details.
- A manuscript draft is checked for invented citations and statistical errors.
- An article is prepared for a particular journal and peer review is simulated.
- A study is packaged for replication and open code.
Not a fit when
- The task is commercial, while the set is released for research.
- You need an autonomous pipeline that writes a whole paper; the full orchestrator is deliberately absent.
- You work outside the social sciences and the skills target journals of that field.
Example request
Run /scholar-init for my interview project and scan the files for personal data before analysis.Limitations
The license is academic and non-commercial: commercial use needs the author's written permission. `scholar-full-paper`, `scholar-grant`, `scholar-teach`, `scholar-book` and `scholar-presentation` are deliberately excluded. The `setup.sh` script changes agent configuration: it creates links and registers hooks in `~/.claude/settings.json`, keeping a backup before editing. It needs `bash`, `python3` and `jq`: without them the guard hook blocks reads. Verify all outputs yourself, since hallucinations and invented references are possible. Reading a file through Claude Code sends its contents to the Anthropic API.
How to disable. Remove the skill and agent links from `~/.claude` and other agent folders, take the hooks out of `~/.claude/settings.json` (a backup sits beside it) and delete the cloned repository.
FAQ
How do I install Open Scholar Skill?
Clone the repository and run `bash setup.sh`. The `--harness claude,codex,zcode` flag limits installation to the agents you need.
Is it suitable for commercial work?
Not without the author's written permission: the license covers academic, educational and non-commercial use.
Does it replace the researcher?
No. The author explicitly asks you to disclose AI use, verify all results and not outsource questions, arguments and interpretations.
How is data protected?
`/scholar-init` scans files locally for personal details and a hook intercepts reads of protected files. A lockdown mode enforces OS-level isolation.
Is there a full pipeline in one command?
There is `scholar-auto-research`, a deterministic skill chain to a verified manuscript. The author does not endorse using it as autonomous research.
Discussion
No questions or reviews yet.
Related
PluginsTen research skills for Claude Code and Codex: figures, conference and journal papers, grants, peer review and code release
Claude Code skills for the academic pipeline: research, paper drafting, peer review and citation checks
A Claude Code skill for scientific literature reviews: an interview, Scopus and OpenAlex search, DOI checks and a final docx file
A Claude Code skill for literature surveys: from a pile of papers to a view of the field via NotebookLM and /survey modes.
A Codex skill that guides a social science paper from topic to citation checks and marks weak spots with risk labels.
A set of three research skills: paper writing, academic Word and PowerPoint documents, scientific computing and figures
README in short
The docs open with the ethics of AI use in research and describe three-layer data protection. Then come a skill table, nine reviewer agents, setting up an article library for `scholar-write` and phone notifications through ntfy or Telegram.