Outline Skill
SkillDev toolsAnalyze paper structure — section balance, word counts, Cochrane-principle compliance
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the Outline 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/outline/SKILL.md and read by ahel’s review.
Review the current paper structure: section balance, word counts, Cochrane-principle compliance, and structural integrity.
Examples
/outline— full structural analysis of main.tex/outline balance— check section length balance only/outline cochrane— Cochrane-principle compliance check only
Workflow
Step 1: Read Current Structure
- Read
main.tex - Extract all
\section{},\subsection{},\subsubsection{}commands - Count lines and estimate words per section (between
%% BEGIN/ENDmarkers) - Verify
%% BEGIN/ENDmarkers are present and match registered keys
Registered section keys: Check the project's CLAUDE.md or guidance/paper-context.md for the section key registry. If neither exists, discover keys by scanning for %% BEGIN: markers in main.tex.
Step 2: Structural Analysis
Section Presence:
- Verify all expected body sections are present (based on registered keys or discovered markers)
- Verify appendices are present if referenced in the body
- Flag any sections that appear in
main.texbut are not in the registered key list
Section Balance:
- Flag sections that are >2x the average body-section length
- Flag sections that are <0.25x the average body-section length
- Introduction target: ~3 pages (Cochrane)
- Conclusion target: ~1--2 paragraphs
Organization (Cochrane Principles):
- Is the main result presented as early as possible?
- Is there anything before the main result that a reader doesn't need?
- Are robustness checks in the appendix (not cluttering the body)?
- Is the literature review AFTER the contribution (not before)?
Step 3: Project-Specific Structural Checks
If guidance/paper-context.md exists, run its registered structural checks. Common checks include:
- Are key figures/tables referenced in the introduction?
- Does the introduction enumerate all main contributions?
- Does the conclusion mention stated deliverables (data, code, packages)?
- Is the paper's terminology used consistently in section headers?
If no paper-context file exists, skip this step and note it in the output.
Step 4: Output
PAPER OUTLINE
=============
Title: [from \title{} command in main.tex]
Current Structure:
1. [Section name] (key: [key]) — N lines, ~M words
2. [Section name] (key: [key]) — N lines, ~M words
...
---
A. [Appendix name] (key: [key]) — N lines, ~M words
...
Total body: ~W words (~P pages at 250 words/page)
Total with appendices: ~W' words
Section Balance:
Average body section: ~N words
Longest: [section] (M words, Xx average)
Shortest: [section] (M words, Xx average)
Cochrane Check:
[x] Main result in first half of paper
[x] Literature review after contribution
[x] Robustness in appendix
[ ] Introduction exceeds 3-page target (currently ~P pages)
Project-Specific Checks:
[x] [check description]
[ ] [any issues found]
(or: "No paper-context.md found — skipping project-specific checks")
Section Markers:
N/N sections have valid %% BEGIN/END markers
[any marker issues]
Signals
- GitHub stars
- 59
- Forks
- 10
- Last commit
- Apr 2026
Advanced
- Catalog kind
- skill
- Gateway key
outline- Source
- github.com/alexander-m-dickerson/ai-asset-pricing