paper

name: paper description: Load context from the paper draft and analysis checklist before working on an analysis. Use when starting work on a paper-related analysis to get relevant context. allowed-tools: Read, Grep, Glob

Paper Analysis Context

Load relevant context from the paper and checklist before starting analysis work.

Usage

/paper [analysis description or notebook name]

Examples:
  /paper 2.01 chemical space
  /paper cluster split
  /paper performance distance

Instructions

1. Read the context files:

   • docs/paper/analysis_checklist.md - analysis tracking (source of truth for status)
   • docs/paper/outline.md - the paper outline

2. Find relevant sections based on the user's query:

   • Search checklist for analysis status and notebook details
   • Search outline for related sections

3. Present a context summary:

   ## Analysis: [name]
   
   ### From Checklist (source of truth)
   - Status: [TODO/IN_PROGRESS/DONE]
   - Expected outputs: [what it should produce]
   - Dependencies: [data or prior analyses needed]
   
   ### From Outline
   - Related sections: [where this fits in the paper]
   - Key context: [relevant excerpts or summary]
   
   ### Ready to Start
   - Inputs available: [yes/no, with paths]
   - Suggested notebook: [path if exists]
   

4. Then stop and let the user drive. Don't start the analysis automatically.

5. After work is done, remind the user:

   • Update checklist status if needed
   • Use /writeup to document findings as a GitHub Issue

Notes

• Workflow: /paper (context) -> analysis -> /writeup (GitHub Issue) -> update checklist
• Source of truth for status: analysis_checklist.md only
• All files evolve - always read fresh, don't assume content
• Keep summaries concise - extract only what's relevant