Paper Outline

Help the researcher design the structure of their paper. A good outline is the skeleton that determines whether the paper's argument flows logically. Produce a section-by-section plan that is specific to this paper, not a generic template.

Step 1: Read Context

Read .papermill/state.md (Read tool) for:

• Thesis: The central claim and novelty (critical -- the outline serves the thesis).
• Prior art: Key references and gaps (these shape the related work section).
• Format: latex, markdown, or rmarkdown (affects section conventions).
• Stage: Current progress.
• Venue: Target venue (affects length, section naming, and emphasis).

If .papermill/state.md does not exist, outlining can still proceed — ask the user to describe the thesis and intended contribution type directly. Suggest running /papermill:init afterward to persist the outline.

Also scan existing paper files (Glob/Read tools). If there is already a partial draft, read it to understand what exists.

Step 2: Determine the Paper Type

Different paper types have different conventional structures. Identify which applies:

• Theoretical/mathematical: Introduction, Preliminaries, Main Results, Proofs, Discussion, Conclusion
• Empirical/experimental: Introduction, Related Work, Method, Experiments, Results, Discussion, Conclusion
• Systems/engineering: Introduction, Background, Design, Implementation, Evaluation, Related Work, Conclusion
• Survey/tutorial: Introduction, Taxonomy/Framework, Section per topic, Discussion, Open Problems
• Short paper/workshop: Compressed -- Introduction, Approach, Results, Conclusion

Ask the user if you are unsure. The structure should match the contribution type.

Step 3: Draft the Outline

For each section, specify:

Field	Description
Section title	The heading as it will appear in the paper
Purpose	What this section accomplishes for the reader (1 sentence)
Key content	Bullet points of what goes here (3-5 items)
Estimated length	Approximate page/paragraph count
Dependencies	What must be established before this section

Example:

3. Preliminaries

Purpose: Establish notation and recall the key definitions the reader needs. Key content:

• Series system model (Definition 2.1)
• Masked failure time observation model
• Exponential lifetime assumption and its implications
• Notation table for symbols used throughout

Estimated length: 1.5 pages Dependencies: None (self-contained background)

Step 4: Check the Narrative Arc

After drafting the outline, verify the story:

1. Does the introduction motivate the problem clearly? The reader should understand why this matters by the end of the first page.
2. Does related work position the contribution? It should make clear what gap this paper fills.
3. Does the technical content flow logically? Each section should build on the previous one.
4. Does the paper deliver on the thesis? The main results section should directly address the claim.
5. Does the conclusion do more than summarize? It should state implications and future work.

Raise any structural issues with the user. Common problems:

• Related work that reads like a list rather than a narrative
• Main results section that is too dense with no intuition
• Missing "so what?" in the conclusion
• Sections that could be merged or reordered for better flow

Step 5: Present and Iterate

Present the complete outline to the user. Ask:

Does this structure capture the argument you want to make? Would you reorder, merge, or split any sections?

Iterate until the user approves.

Step 6: Update State File

Once approved, update .papermill/state.md (Edit tool):

• Set stage to outlining (or drafting if progressing).
• Append the outline to the markdown body under a ## Outline heading.

Append a timestamped note documenting the outline creation.

Step 7: Suggest Next Steps

Based on what the outline reveals, suggest the most relevant next step:

• If the outline exposed that the thesis is unclear or too broad → "The outlining process suggests the thesis may need sharpening. Consider running /papermill:thesis to refine the claim before drafting."
• If the outline includes a related work section with few references → /papermill:prior-art
• If the outline includes experiment/simulation sections → /papermill:experiment or /papermill:simulation
• If the outline includes proof sections → /papermill:proof
• Otherwise → "Begin writing with the section you feel most confident about. Many authors start with the method/results, not the introduction."