agentleFS
yyyy/claude-scholar/skills/presubmit-checks/SKILL.md

presubmit-checks

Skill39 starsChanged 8 months agoMITView on GitHub

Check a LaTeX paper before submission or circulation. Verify references, compilation, LaTeX hygiene, figure quality, affiliations, acknowledgements, and code or data availability, then return one severity-ranked report.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 2 are distinctive, and 4 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.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use a manuscript path supplied by the user.

    line 13
  2. Distinctive:

    Use the project's documented build command.

    line 15
  3. Very specific:

    Use the check-refs skill to verify cited works.

    line 21
  4. Very specific:

    Use the latex-cleanup skill for source-level checks.

    line 25
  5. Very specific:

    Use the critique-figures skill on the manuscript's rendered figures.

    line 36
  6. Distinctive:

    Every author and affiliation.

    line 44

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Check a paper before submission

Inspect the paper without modifying it. Run independent checks concurrently when practical; the coordination method and number of agents may vary.

Establish scope

Use a manuscript path supplied by the user. Otherwise, infer the main file from project configuration, build files, and the TeX document tree. Follow \input and \include dependencies and resolve figure and bibliography paths from the main document.

Use the project's documented build command. If none is documented and a Makefile exists, use its default target. Record the command and working directory so failures are reproducible.

Run the checks

References

Use the check-refs skill to verify cited works. Report retracted citations as blockers, distinguish unverified entries from nonexistent works, and note missing DOI or URL metadata separately.

LaTeX and build

Use the latex-cleanup skill for source-level checks. Build the paper and inspect the log for:

  • Compilation errors and missing files.
  • Undefined references or citations.
  • Draft artifacts and draft mode.
  • Overfull boxes that visibly damage layout.

Continue the other checks if the build fails.

Figures

Use the critique-figures skill on the manuscript's rendered figures. Incorporate its findings on format, effective resolution, legibility, accessibility, visual encoding, panel labels, and statistical presentation.

Carry every blocker identified by critique-figures into the unified report. In particular, raster-only plots, diagrams, and line art remain submission blockers; apply that skill's rules for pixel-native and mixed figures.

Front matter

Extract enough information for the user to verify:

  • Every author and affiliation. Flag missing affiliations and apparently incomplete department, institution, or country information.
  • The acknowledgements, including every funding source, grant number, and named person thanked. Flag a missing acknowledgements section.
  • Code and data availability statements. Flag a missing statement as a blocker only when the target venue requires one; otherwise report it as a warning.

Report

Return one concise report organized by severity:

  1. Blockers: compilation errors, missing figures, undefined references or citations, retracted citations, raster-only plots or line art, and missing venue-required availability statements.
  2. Warnings: unverified references, draft artifacts, consequential overfull boxes, incomplete affiliations, material figure-quality findings, and other likely submission problems.
  3. User verification: list the extracted affiliations, acknowledgements, funding details, and availability statements for completeness checks.
  4. Suggestions: missing DOI or URL metadata and non-blocking typography or style improvements.

Give file and line references or log excerpts for each actionable finding. Group repeated warnings and state which checks could not be completed. Do not modify files unless the user approves a separate fix pass. Use critique-manuscript only when the user also requests a content-level review.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.