ai-paper-check
Skill0 starsChanged 29 days agoMITView on GitHub
Check AI research manuscripts for notation and acronym consistency, figures and tables, reproducibility materials, citation and numerical accuracy, LaTeX builds, and venue requirements. Use for technical audits, compilation repair, and submission checks. Scientific argument review belongs to ai-paper-review; prose rewriting belongs to ai-paper-writing.
In this repository2 other files it gives its agents.
- ai-paper-reviewSkill
- ai-paper-writingSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Read the shared principles.
- Distinctive: line 29
Use the findings contract.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AI Paper Check
Read the shared principles. Verify factual and technical properties against supplied artifacts or authoritative sources. Check the requested scope; a bibliography audit, build repair, and submission check are distinct tasks.
Select the relevant checks
| Task | Guidance |
| Citation metadata, cited claims, BibTeX, numerical or terminology consistency | Evidence and citation checks |
| Symbol definitions, consistent notation, and acronym expansion at first use | Notation and abbreviations |
| Figure and table references, captions, axes, legends, values, and highlighting | Figures and tables |
| Experimental protocols, software, resources, prompts, and reproduction instructions | Reproducibility materials |
| LaTeX compilation, labels, PDF layout, anonymity, page limits, submission package | Build and submission checks |
| Main/appendix correspondence and proof placement | Appendix organization |
Establish the source of truth before correcting anything: result files or logs, manuscript equations, authoritative publication records, or official venue instructions. Reuse existing verified records when applicable. The author's paper plan records intent; it does not override contradictory evidence.
Load the detailed checklists that match the request. For a full technical manuscript audit, include notation, figures/tables, and applicable reproducibility materials alongside evidence and build checks. A targeted acronym or figure check does not require auditing the entire submission package.
Keep a check-only task read-only. Fix identified defects when repair is requested or implied, preserving valid citation keys, values, scientific meaning, and the existing build engine. Resolve ambiguities from available evidence; when an authoritative value is unavailable, report the conflict rather than selecting a plausible value. Continue independent checks when one source or tool is unavailable.
Report verified scope
Use the findings contract. State the files and checks covered, evidence consulted, and remaining blockers for the requested readiness decision.
A successful build does not verify scientific claims. Matching BibTeX metadata does not establish that a source supports the cited sentence. Passing technical checks does not certify novelty, acceptance, or the full scientific argument. Use ai-paper-review only for a requested or materially needed argument assessment; do not start an automatic review-and-rewrite cycle.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.