
Paper Plan Outliner
Turns review conclusions and experiment results into a section-by-section paper outline for a target venue.
- Stars
- 18,482
- Licence
- MIT
- Files
- 1 file
What it does
When you ask the agent to plan a research paper, this skill makes it gather your claims, evidence and reviewer feedback from project files and build a structured outline. The outline covers each section's length, key citations, hero figure and claims-evidence table, sized to a chosen venue's page limit (ICLR, NeurIPS, IEEE and others). It is designed to be used with a separate Claude Code reviewer, so the review step is handled by a different model family than the one writing the plan.
Great for
- Researchers preparing a conference or journal submission who already have experiment results and review notes
- Students who want a clear structure before drafting a paper
- Teams using the Auto-claude-code-research-in-sleep workflow who want the paper plan to follow earlier review stages
Skip it if
- People with no experiment results or research claims yet, since the skill asks for a 3-5 sentence description instead and produces a weaker plan
- Anyone expecting it to write the full paper text; it produces an outline and planning notes, not finished prose
Try asking
- “Write a paper outline for my ICLR submission based on AUTO_REVIEW.md and my results figures”
- “Plan the paper for NeurIPS using my CLAIMS_FROM_RESULTS.md”
- “Create a section-by-section plan for an IEEE journal paper with a 12-14 page limit”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
paper-plan.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip paper-plan.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip paper-plan.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Run result-to-claim first so CLAIMS_FROM_RESULTS.md exists; the plan is more reliable when claims are already adjudicated
- State your target venue explicitly, since page budgets differ widely between venues
- Check the venue's current call for papers, because the skill notes page limits may change
What's inside
- SKILL.md
Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.