Submission Prep Skill

SkillDev tools

Pre-submission checklist for JF, RFS, JFE, or other target journals

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Submission Prep Skill skill

What this skill tells your AI

The instructions your AI receives, as published by alexander-m-dickerson/ai-asset-pricing in .claude/skills/submission-prep/SKILL.md and read by ahel’s review.

Run a pre-submission checklist tailored to the target journal. Catches common rejection-worthy issues before you submit.

Examples

  • /submission-prep -- run full checklist (uses target journal from project's CLAUDE.md)
  • /submission-prep JF -- run checklist with JF-specific requirements
  • /submission-prep --strict -- flag warnings as failures

Workflow

Step 1: Load Target Journal

Read the project's CLAUDE.md for the target journal. If not specified, ask the user. Common targets: JF, RFS, JFE, RoF, JFQA, MS.

Step 2: Run Checks

2a. Document Structure

  • Abstract present and within word limit (100--150 words for JF/RFS/JFE)
  • Title length reasonable (<15 words recommended)
  • All expected body sections present (check project's CLAUDE.md for section list)
  • Page count within typical range (40--60 pages including appendices for top-3)

2b. Content Completeness

  • No [HUMAN EDIT REQUIRED] tags remaining in any .tex file
  • No TODO, FIXME, XXX comments in .tex
  • No \lipsum or placeholder text
  • No commented-out sections that should be removed or restored

2c. Terminology Compliance

  • Project-specific terminology used consistently (check project's CLAUDE.md)
  • No banned words from academic-writing.md Section 1

2d. Tables and Figures

  • All tables have self-contained captions (sample period, units, variable definitions)
  • All figures have self-contained captions with axis labels
  • All tables and figures are referenced in the text (\ref check)
  • Numbers use 2--3 significant digits (not computer output)
  • Tables use booktabs style (\toprule/\midrule/\bottomrule, no vertical lines)
  • Figures use vector format (PDF) where possible

2e. Bibliography

  • All \cite{} keys resolve to .bib entries
  • No unused .bib entries
  • No duplicate .bib entries
  • Consistent BibTeX format (all @article have journal, year, volume, pages)
  • Proper nouns protected in titles: {CAPM}, {U.S.}, {NYSE}, etc.

2f. Cross-References

  • All \ref{} and \eqref{} resolve to valid labels
  • Equations use \eqref not \ref
  • Non-breaking spaces used (Table~\ref, Eq.~\eqref, Figure~\ref)

2g. Writing Quality

  • No banned words (run /style-check if not done recently)
  • No AI-marker words (see academic-writing.md for full list)
  • No em-dashes (---) in prose
  • No structural AI tells: naked "this", adverb openers, "Together, these results..."
  • Soft-ban words within limits: "highlights" (max 2/paper), "insights" (max 1/paper)
  • No hedge words: somewhat, quite, very (intensifier), arguably, perhaps
  • No previewing: "as we show below", "we will show", "Recall from"
  • Active voice throughout
  • Specific numbers for quantitative claims
  • First sentence is concrete, not throat-clearing

2h. Compilation

  • Clean compile with no errors
  • Acceptable warning count (flag if >10 non-font warnings)
  • PDF renders correctly (no missing figures, no ?? references)

2i. Journal-Specific Requirements

  • Author information complete (names, affiliations, emails)
  • Acknowledgments section present
  • JEL classification codes present
  • Keywords present (if required by target journal)
  • Data availability statement present (increasingly required)

2j. Key Results Verification

  • Key claims in the project's CLAUDE.md match the numbers in tables/text
  • Sample period stated consistently across abstract, introduction, and table captions
  • Main results discussed in both introduction and results section

Step 3: Output

SUBMISSION PREP CHECKLIST
=========================

Target: [journal]
Paper: [title from project CLAUDE.md]

PASS: N checks passed
FAIL: M checks failed
WARN: K warnings

--- FAILURES ---
[ ] Line X: [HUMAN EDIT REQUIRED] tag found in section Y
[ ] Table Z: caption missing sample period
[ ] \cite{smith2023}: key not in .bib

--- WARNINGS ---
[!] Abstract is 168 words (target: 100-150)
[!] 14 compilation warnings (non-font)
[!] Page count: 65 pages (typical: 40-60)

--- PASSED ---
[x] All body sections present
[x] All figures referenced in text
[x] Clean compile (0 errors)
[x] Terminology compliant
[etc.]

RECOMMENDATION: [READY TO SUBMIT / FIX N ISSUES FIRST]

Signals

GitHub stars
59
Forks
10
Last commit
Apr 2026
Advanced
Catalog kind
skill
Gateway key
submission-prep
Source
github.com/alexander-m-dickerson/ai-asset-pricing