Override for Codex users who want Gemini, not a second Codex agent, to act as the reviewer. Install this package after skills/skills-codex/*.

Workflow 3: Paper Writing Pipeline

Gemini overlay assurance: review_independence: cross-family and acceptance_status: accepted.

Orchestrate a complete paper writing workflow for: $ARGUMENTS

Overview

This skill chains five sub-skills into a single automated pipeline:

/paper-plan → /paper-figure → /paper-write → /paper-compile → /auto-paper-improvement-loop
  (outline)     (plots)        (LaTeX)        (build PDF)       (review & polish ×2)

Each phase builds on the previous one's output. The final deliverable is a polished, reviewed paper/ directory with LaTeX source and compiled PDF.

Constants

• VENUE = ICLR — Target venue. Options: ICLR, NeurIPS, ICML. Affects style file, page limit, citation format.
• MAX_IMPROVEMENT_ROUNDS = 2 — Number of review→fix→recompile rounds in the improvement loop.
• REVIEWER_MODEL = gemini-review — Gemini reviewer invoked through the local gemini-review MCP bridge for plan review, figure review, writing review, and the improvement loop.
• AUTO_PROCEED = true — Auto-continue between phases. Set false to pause and wait for user approval after each phase.
• HUMAN_CHECKPOINT = false — When true, the improvement loop (Phase 5) pauses after each round's review to let you see the score and provide custom modification instructions. When false (default), the loop runs fully autonomously. Passed through to /auto-paper-improvement-loop.

Override inline: /paper-writing "NARRATIVE_REPORT.md" — venue: NeurIPS, human checkpoint: true

Inputs

This pipeline accepts one of:

1. NARRATIVE_REPORT.md (best) — structured research narrative with claims, experiments, results, figures
2. Research direction + experiment results — the skill will help draft the narrative first
3. Existing PAPER_PLAN.md — skip Phase 1, start from Phase 2

The more detailed the input (especially figure descriptions and quantitative results), the better the output.

Pipeline

Phase 1: Paper Plan

Invoke /paper-plan to create the structural outline:

/paper-plan "$ARGUMENTS"

What this does:

• Parse NARRATIVE_REPORT.md for claims, evidence, and figure descriptions
• Build a Claims-Evidence Matrix — every claim maps to evidence, every experiment supports a claim
• Design section structure (5-8 sections depending on paper type)
• Plan figure/table placement with data sources
• Scaffold citation structure
• Gemini reviews the plan for completeness via the /paper-plan overlay

Output: PAPER_PLAN.md with section plan, figure plan, citation scaffolding.

Checkpoint: Present the plan summary to the user.

📐 Paper plan complete:
- Title: [proposed title]
- Sections: [N] ([list])
- Figures: [N] auto-generated + [M] manual
- Target: [VENUE], [PAGE_LIMIT] pages

Shall I proceed with figure generation?

• User approves (or AUTO_PROCEED=true) → proceed to Phase 2.
• User requests changes → adjust plan and re-present.

Phase 2: Figure Generation

Invoke /paper-figure to generate data-driven plots and tables:

/paper-figure "PAPER_PLAN.md"

What this does:

• Read figure plan from PAPER_PLAN.md
• Generate matplotlib/seaborn plots from JSON/CSV data
• Generate LaTeX comparison tables
• Create figures/latex_includes.tex for easy insertion
• Gemini reviews figure quality and captions via the /paper-figure overlay

Output: figures/ directory with PDFs, generation scripts, and LaTeX snippets.

Phase 2b: AI Illustration Generation (when illustration: true)

Skip this step entirely if illustration is not set or is false.

If the paper plan includes architecture diagrams, pipeline figures, or method illustrations, invoke /paper-illustration:

/paper-illustration "[method description from PAPER_PLAN.md or NARRATIVE_REPORT.md]"

What this does:

• Codex plans the layout → Gemini optimizes → Nano Banana Pro renders → Codex reviews (score ≥ 9)
• Output: figures/ai_generated/*.png — publication-quality method diagrams
• Requires GEMINI_API_KEY environment variable

Without illustration: true: Architecture diagrams must still be created manually (draw.io, Figma, TikZ) and placed in figures/ before proceeding — same as before.

Checkpoint: List generated vs manual figures.

📊 Figures complete:
- Data plots (auto): [list]
- AI illustrations (auto): [list, if illustration: true]
- Manual (need your input): [list]
- LaTeX snippets: figures/latex_includes.tex

[If manual figures needed]: Please add them to figures/ before I proceed.
[If all auto]: Shall I proceed with LaTeX writing?

Writing invariant (every drafting and revision step): calibrate each claim to its evidence and state it directly; generic caveats live in the Limitations section only; writing instructions are never manuscript content ("do not mention X" means omit X, not "we do not address X"); tone edits never change what the paper knows; the paper is a launch, not a progress report — organize around the strongest advantage, give every experiment an argumentative duty, keep unfavorable numbers in the tables and explain them as tradeoffs only where the evidence supports that, never narrating defeats. /paper-write carries the full CONFIDENT PROSE, HONEST LIMITS contract.

Phase 3: LaTeX Writing

Invoke /paper-write to generate section-by-section LaTeX:

/paper-write "PAPER_PLAN.md"

What this does:

• Write each section following the plan, with proper LaTeX formatting
• Insert figure/table references from figures/latex_includes.tex
• Build references.bib from citation scaffolding
• Clean stale files from previous section structures
• Automated bib cleaning (remove uncited entries)
• De-AI polish (remove "delve", "pivotal", "landscape"...)
• Gemini reviews each section for quality via the /paper-write overlay

Output: paper/ directory with main.tex, sections/*.tex, references.bib, math_commands.tex.

Checkpoint: Report section completion.

✍️ LaTeX writing complete:
- Sections: [N] written ([list])
- Citations: [N] unique keys in references.bib
- Stale files cleaned: [list, if any]

Shall I proceed with compilation?

Phase 4: Compilation

Invoke /paper-compile to build the PDF:

/paper-compile "paper/"

What this does:

• latexmk -pdf with automatic multi-pass compilation
• Auto-fix common errors (missing packages, undefined refs, BibTeX syntax)
• Up to 3 compilation attempts
• Post-compilation checks: undefined refs, page count, font embedding
• Precise page verification via pdftotext
• Stale file detection

Output: paper/main.pdf

Checkpoint: Report compilation results.

🔨 Compilation complete:
- Status: SUCCESS
- Pages: [X] (main body) + [Y] (references) + [Z] (appendix)
- Within page limit: YES/NO
- Undefined references: 0
- Undefined citations: 0

Shall I proceed with the improvement loop?

Phase 5: Auto Improvement Loop

Invoke /auto-paper-improvement-loop to polish the paper:

/auto-paper-improvement-loop "paper/"

What this does (2 rounds):

Round 1: Gemini reviews the full paper → identifies CRITICAL/MAJOR/MINOR issues → Codex implements fixes → recompile → save main_round1.pdf

Round 2: Gemini re-reviews the recompiled draft cold (fresh review — no fix summaries, no conversation carry-over) → identifies remaining issues → Codex implements fixes → recompile → save main_round2.pdf

Typical improvements (calibration cuts both ways):

• Fix assumption-model mismatches
• Narrow genuine overclaims to the supported scope — and state supported claims directly, removing redundant hedges
• Consolidate scattered generic caveats into Limitations
• Add missing interpretations and notation
• Make Limitations more specific (only when a material limit is missing — never pad)
• Add theory-aligned experiments if needed

Output: Three PDFs for comparison + PAPER_IMPROVEMENT_LOG.md.

Format check (included in improvement loop Step 8): After final recompilation, auto-detect and fix overfull hboxes (content exceeding margins), verify page count vs venue limit, and ensure compact formatting. Any overfull > 10pt is fixed before generating the final PDF.

Phase 5.9: Integrity Forensics (opt-in — Codex executor rules apply)

Same contract as the base Codex pack (skills-codex/paper-writing Phase 5.9, which this overlay does not change): the executor here is Codex, so only upstream's deterministic-only slice of /integrity-forensics is runnable (numeric core + rules-only reporter — it can flag, it can never say CLEAN), OFF unless the CURRENT $ARGUMENTS contains — self_forensics: true. If opted in: run the launcher on paper/ (absolute path), then before the Final Report require python3 "$GATE_HELPER" fresh --paper-dir paper/ --anti-ar-commit "$ANTI_AR_COMMIT" exit 0 — an opted-in run without a fresh gate is incomplete, not skippable; gate BLOCK refuses the Final Report.

Phase 6: Final Report

# Paper Writing Pipeline Report

**Input**: [NARRATIVE_REPORT.md or topic]
**Venue**: [ICLR/NeurIPS/ICML]
**Forensics**: [NO_NEW_BLOCKER | WARN: <n> open obligations (listed below) | BLOCK | n/a (opt-in, not requested)]
**Date**: [today]

## Pipeline Summary

| Phase | Status | Output |
|-------|--------|--------|
| 1. Paper Plan | ✅ | PAPER_PLAN.md |
| 2. Figures | ✅ | figures/ ([N] auto + [M] manual) |
| 3. LaTeX Writing | ✅ | paper/sections/*.tex ([N] sections, [M] citations) |
| 4. Compilation | ✅ | paper/main.pdf ([X] pages) |
| 5. Improvement | ✅ | [score0]/10 → [score2]/10 |

## Improvement Scores
| Round | Score | Key Changes |
|-------|-------|-------------|
| Round 0 | X/10 | Baseline |
| Round 1 | Y/10 | [summary] |
| Round 2 | Z/10 | [summary] |

## Deliverables
- paper/main.pdf — Final polished paper
- paper/main_round0_original.pdf — Before improvement
- paper/main_round1.pdf — After round 1
- paper/main_round2.pdf — After round 2
- paper/PAPER_IMPROVEMENT_LOG.md — Full review log

## Remaining Issues (if any)
- [items from final review that weren't addressed]

## Next Steps
- [ ] Visual inspection of PDF
- [ ] Add any missing manual figures
- [ ] Submit to [venue] via OpenReview / CMT / HotCRP

Output Protocols

Follow these shared protocols for all output files:

• Output Versioning Protocol — write timestamped file first, then copy to fixed name
• Output Manifest Protocol — log every output to MANIFEST.md
• Output Language Protocol — respect the project's language setting
• Note: paper content is always written in English regardless of project language setting.

Key Rules

• Large file handling: If the Write tool fails due to file size, immediately retry using Bash (cat << 'EOF' > file) to write in chunks. Do NOT ask the user for permission — just do it silently.

• Don't skip phases. Each phase builds on the previous one — skipping leads to errors.

• Checkpoint between phases when AUTO_PROCEED=false. Present results and wait for approval.

• Manual figures first. If the paper needs architecture diagrams or qualitative results, the user must provide them before Phase 3.

• Compilation must succeed before entering the improvement loop. Fix all errors first.

• Preserve all PDFs. The user needs round0/round1/round2 for comparison.

• Document everything. The pipeline report should be self-contained.

• Respect page limits. If the paper exceeds the venue limit, suggest specific cuts before the improvement loop.

Composing with Other Workflows

/idea-discovery "direction"         ← Workflow 1: find ideas
implement                           ← write code
/run-experiment                     ← deploy experiments
/auto-review-loop "paper topic"     ← Workflow 2: iterate research
/paper-writing "NARRATIVE_REPORT.md"  ← Workflow 3: you are here
                                         submit! 🎉

Or use /research-pipeline for the Workflow 1+2 end-to-end flow,
then /paper-writing for the final writing step.

Typical Timeline

Total: ~45-90 min for a full paper from narrative report to polished PDF.