paper-writing

name	paper-writing
description	Workflow 3: evidence-gated paper writing pipeline. Requires audited formal experiment evidence and an accepted paper plan before manuscript drafting. Orchestrates paper-plan → paper-figure → figure-spec/paper-illustration/mermaid-diagram → paper-write → paper-compile → auto-paper-improvement-loop only after the manuscript entry gate passes. If only smoke, pilot, toy, or validation results exist, stop and produce next actions or a gap report. Use when user says "写论文全流程", "write paper pipeline", "从报告到PDF", "paper writing", or wants complete paper generation from audited evidence.
argument-hint	[narrative-report-path-or-topic] [— style-ref: <source>]
allowed-tools	Bash(*), Read, Write, Edit, Grep, Glob, Skill, mcp__codex__codex, mcp__codex__codex-reply

Workflow 3: Paper Writing Pipeline

Capability Routing

This is a first-layer entry skill. Keep it loaded as the user-facing route; when a request needs a specialized capability below, resolve the debuffer repo root from .debuffer_skills/installed-skills-codex.txt (repo_root) when available, read the referenced library SKILL.md, then follow that skill. Do not copy the whole library skill into this file.

• /paper-plan: read ../library/paper/paper-plan/SKILL.md.
• /paper-write: read ../library/paper/paper-write/SKILL.md.
• /paper-compile: read ../library/paper/paper-compile/SKILL.md.
• /overleaf-package: read ../library/paper/overleaf-package/SKILL.md.
• /artifact-package-splitter: read ../library/paper/artifact-package-splitter/SKILL.md.
• /paper-claim-audit: read ../library/review/paper-claim-audit/SKILL.md.
• /citation-audit: read ../library/review/citation-audit/SKILL.md.
• /paper-math-consistency-audit: read ../library/review/paper-math-consistency-audit/SKILL.md.
• /final-experiment-curator: read ../library/review/final-experiment-curator/SKILL.md.

Orchestrate an evidence-gated paper writing workflow for: $ARGUMENTS

Customized Pack Defaults

Read ../shared-references/lightweight-research-pack.md, ../shared-references/project-guide-protocol.md, and ../shared-references/venue-profiles.md, plus ../shared-references/icde-yu-memory-paper-structure.md before starting. Defaults:

• AUTO_PROCEED = false, HUMAN_CHECKPOINT = true, MAX_IMPROVEMENT_ROUNDS = 1, REVIEW_MODE = prompt-only.
• MANUSCRIPT_ENTRY_GATE = required: formal runs and evidence audit must pass before creating LaTeX prose or updating PROJECT_STATUS.md to manuscript.
• VENUE supports ICLR, AAAI, JMLR, TPAMI, NeurIPS, ICML, and existing legacy venue labels. Apply the venue profile before outline and audit prompts.
• Prefer compact project memory (PROJECT_STATUS.md, PROJECT_BRIEF.md, PAPER_GUIDE.md, EVIDENCE_LEDGER.md, EXPERIMENT_PROTOCOL.md, findings.md, EXPERIMENT_LOG.md, CLAIMS_FROM_RESULTS.md) and existing PAPER_PLAN.md before generating or reading a large NARRATIVE_REPORT.md.
• Update PROJECT_STATUS.md to manuscript only after the manuscript entry gate passes and paper drafting starts; update it to submission only after venue formatting and required audits/checks pass.
• Create or refresh PAPER_GUIDE.md at the manuscript gate by compacting the project guide, experiment protocol, evidence ledger, and accepted paper plan. Do not duplicate full audit reports inside the guide.
• Do not call reviewer MCP/API backends by default. Generate prompts under review-prompts/ for paper-plan, claim, citation, and improvement review, then wait for pasted feedback from a separate conversation.
• The final pipeline report should be concise. Keep detailed audit outputs in their existing files and summarize only blockers, accepted changes, and next actions.

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.

In this hybrid pack, the pipeline itself is unchanged, but paper-plan, paper-write, and paper-compile use shared references for stronger story framing, prose guidance, and submission-time checks. It also benefits from /figure-table-audit and /experiment-writeup-audit before finalization.

Default manuscript package target: the ICDE_YU_Memory modular Overleaf style or better. For new IEEE/Overleaf starts, prefer thin main.tex, Content/ section files, Figure/ assets, and IEEE.bib; preserve an existing sections/ / figures/ / references.bib convention only for existing projects or non-IEEE templates. Use the shared reference above for section topology, RQ-organized experiments, benchmark-construction sections, and discussion/related-work placement.

Constants

• VENUE = ICLR — Target venue. Options: ICLR, NeurIPS, ICML, CVPR, ACL, AAAI, JMLR, TPAMI, ACM, IEEE_JOURNAL (IEEE Transactions / Letters), IEEE_CONF (IEEE conferences). Affects style file, page limit, citation format.
• MAX_IMPROVEMENT_ROUNDS = 1 — Number of review→fix→recompile rounds in the improvement loop.
• REVIEWER_MODEL = gpt-5.5 — Model used via Codex MCP for plan review, figure review, writing review, and improvement loop.
• AUTO_PROCEED = false — Pause between phases unless the user explicitly opts into auto-continue.
• HUMAN_CHECKPOINT = true — Pause after each improvement review so the user can inspect the score and provide custom modification instructions. Passed through to /auto-paper-improvement-loop.
• ILLUSTRATION = figurespec — Architecture/illustration generator for Phase 2b: figurespec (default, deterministic JSON→SVG via /figure-spec, best for architecture/workflow/topology), gemini (AI-generated via /paper-illustration, best for qualitative method illustrations; needs GEMINI_API_KEY), codex-image2 (AI-generated via /paper-illustration-image2 through the local Codex native image bridge — no external API key, uses your ChatGPT Plus/Pro quota; experimental), mermaid (Mermaid syntax via /mermaid-diagram, free, best for flowcharts), or false (skip Phase 2b, manual only).

Override inline: /paper-writing "NARRATIVE_REPORT.md" — venue: NeurIPS, illustration: gemini, human checkpoint: true IEEE example: /paper-writing "NARRATIVE_REPORT.md" — venue: IEEE_JOURNAL

Inputs

This pipeline accepts one of:

1. Accepted docs/paper/PAPER_PLAN.md plus audited formal evidence — best.
2. NARRATIVE_REPORT.md plus audited formal evidence — allowed only if the entry gate passes.
3. Research direction + validation results — not enough for manuscript drafting; produce docs/project/NEXT_ACTIONS.md or docs/paper/GAP_REPORT.md and stop.

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

Optional: Style reference (— style-ref: <source>, opt-in)

Lets the user steer structural style (section ordering, theorem density, sentence cadence, figure density, bibliography style) of the generated paper toward a reference paper they admire. Default OFF — when the user does not pass — style-ref, do nothing differently from before.

When — style-ref: <source> is in $ARGUMENTS, run the helper FIRST, before Phase 1 (paper-plan):

# Resolve $STYLE_HELPER via the canonical strict-safe chain (see
# shared-references/integration-contract.md §2). Policy A — gate:
# unresolved helper means --style-ref cannot be satisfied, so abort.
cd "$(git rev-parse --show-toplevel 2>/dev/null || pwd)" || exit 1
if [ -z "${ARIS_REPO:-}" ] && [ -f .debuffer_skills/installed-skills.txt ]; then
    ARIS_REPO=$(awk -F'\t' '$1=="repo_root"{print $2; exit}' .debuffer_skills/installed-skills.txt 2>/dev/null) || true
fi
STYLE_HELPER=".debuffer_skills/tools/extract_paper_style.py"
[ -f "$STYLE_HELPER" ] || STYLE_HELPER="tools/extract_paper_style.py"
[ -f "$STYLE_HELPER" ] || { [ -n "${ARIS_REPO:-}" ] && STYLE_HELPER="$ARIS_REPO/tools/extract_paper_style.py"; }
[ -f "$STYLE_HELPER" ] || {
  echo "ERROR: extract_paper_style.py not resolved at .debuffer_skills/tools/, tools/, or \$ARIS_REPO/tools/." >&2
  echo "       Fix: rerun bash tools/install_aris.sh, export ARIS_REPO, or copy the helper to tools/." >&2
  echo "       --style-ref cannot be satisfied; aborting." >&2
  exit 1
}
STYLE_STATUS=0
CACHE=$(python3 "$STYLE_HELPER" --source "<source>") || STYLE_STATUS=$?
case "$STYLE_STATUS" in
  0) ;;                                       # share $CACHE/style_profile.md with downstream WRITER phases only
  2) echo "warning: style-ref skipped (missing optional dep)" >&2 ;;
  3) echo "error: --style-ref source failed; aborting pipeline" >&2 ; exit 1 ;;
  *) echo "error: helper failed unexpectedly; aborting pipeline" >&2 ; exit 1 ;;
esac

Then forward — style-ref: <source> only to the writer-side sub-skills:

• /paper-plan (Phase 1) — outline structure
• /paper-write (Phase 3) — section-by-section prose
• /paper-illustration (Phase 2b) — figure structural matching, optional

Sources accepted: local TeX dir / file, local PDF, arXiv id, http(s) URL. Overleaf URLs/IDs are rejected — clone via /overleaf-sync setup <id> first and pass the local clone path.

Strict rules (full contract in tools/extract_paper_style.py docstring):

• Use style_profile.md as structural guidance only. Match section-count tendency, theorem density, caption-length distribution, sentence cadence, math display ratio, citation style.
• Never copy prose, claims, examples, or terminology from anything reachable through the cache.
• Never pass — style-ref (or the cache contents) to reviewer / auditor sub-skills — Phase 4.5 (/proof-checker), Phase 4.6 / 5.7 (/paper-math-consistency-audit), Phase 4.7 / 5.5 (/paper-claim-audit), Phase 5 (/auto-paper-improvement-loop reviewer), Phase 5.8 (/citation-audit) MUST run on the artifact alone. Cross-model review independence (../shared-references/reviewer-independence.md).

Pipeline

Phase 0: Manuscript Entry Gate

Before assurance setup, inspect PROJECT_STATUS.md, docs/project/BLUEPRINT_GATE.md, docs/paper/PAPER_PLAN.md, docs/evidence/EVIDENCE_LEDGER.md, CLAIMS_FROM_RESULTS.md, docs/experiments/EXPERIMENT_LOG.md, and raw result folders when present.

Proceed only if all are true:

1. Formal baseline/main/required ablation runs exist for paper-level claims.
2. Raw evidence is auditable: run folders, metrics, configs/resolved configs, seeds, logs/metadata, and result summaries are named.
3. docs/evidence/EVIDENCE_LEDGER.md, CLAIMS_FROM_RESULTS.md, or an equivalent audit maps claims to raw evidence and unresolved gaps.
4. docs/paper/PAPER_PLAN.md exists and its next-step block says the manuscript entry gate passed, or the user explicitly asks to create the plan first and the evidence gate is already passed.
5. The paper can state a stable one-sentence core contribution, with separately named research question, main finding, and significance. If any of these is still changing, return to evidence/plan work instead of drafting prose.
6. The accepted plan contains a Claim–Evidence–Figure Storyboard: every primary figure/table answers one question, maps to a raw-result locator, names the claim it supports, and records the strongest permitted conclusion plus any missing evidence or analysis.
7. The target venue range is known and unresolved user decisions are explicitly listed rather than silently decided in the draft.

If any check fails, stop before Phase 1. Do not create paper/, LaTeX section files, or docs/paper/PAPER_GUIDE.md; do not update PROJECT_STATUS.md to manuscript. Write the smallest useful docs/project/NEXT_ACTIONS.md or docs/paper/GAP_REPORT.md naming the missing formal experiments/audits and the next gate.

Smoke, pilot, toy, or validation runs never satisfy this gate.

This is a readiness gate, not a calendar promise. Once it passes, compose the first draft in evidence order (storyboard → Methods → Results → Discussion → Introduction → Abstract/title) while preserving the plan's conventional final section order. /paper-plan owns the storyboard contract; /paper-write owns the composition order and section-level evidence discipline.

Phase 0.5: Assurance Setup

Before assurance setup, read PROJECT_STATUS.md and PAPER_GUIDE.md when present. If PAPER_GUIDE.md is missing but the project has stable PROJECT_GUIDE.md, EXPERIMENT_PROTOCOL.md, EVIDENCE_LEDGER.md, or an accepted PAPER_PLAN.md, compact those into PAPER_GUIDE.md instead of creating another long narrative report.

Resolve the active assurance level and persist it so Phase 6's external verifier reads the same value. Run once at pipeline start, before Phase 1.

Resolution order (first match wins):

1. Explicit — assurance: draft | submission in $ARGUMENTS
2. Derived from — effort:
   • lite / balanced → draft (default, zero change from current behavior)
   • max / beast → submission
3. Default: draft

Action:

mkdir -p paper/.debuffer_skills
echo "<resolved-level>" > paper/.debuffer_skills/assurance.txt   # draft or submission

What each level does downstream:

• draft — Existing behavior. Audits run only when their content detector matches (Phase 4.5 / 4.6 / 4.7 / 5.5 / 5.7 / 5.8). Missing artifacts are non-blocking. Silent-skip allowed.
• submission — The mandatory audits (proof-checker, paper-math-consistency-audit, paper-claim-audit, kill-argument, and citation-audit) are treated as load-bearing gates. Each sub-audit must emit its JSON artifact (PASS / WARN / FAIL / NOT_APPLICABLE / BLOCKED / ERROR) — never silent-skip. Phase 6 runs verify_paper_audits.sh (canonical name; resolved per shared-references/integration-contract.md §2); a non-zero exit blocks the Final Report.

Escape hatch: a user wanting the old "beast = depth-only, no audit gate" can pass — effort: beast, assurance: draft explicitly. Legal but discouraged for actual submissions. See shared-references/assurance-contract.md for the full contract.

Announce the resolved level in-line before Phase 1:

📋 Assurance: <level> (derived from effort: <effort>)
   <either "current behavior, no audit gate" OR "mandatory audits gated by verify_paper_audits.sh (resolved per integration-contract §2)">

Phase 1: Paper Plan

Invoke /paper-plan to create the structural outline:

/paper-plan "$ARGUMENTS"

If — style-ref: <source> was passed in $ARGUMENTS and the helper succeeded above, append — style-ref: <source> to the invocation: /paper-plan "<topic> — style-ref: <source>". (Writer-side phase — forwarding is allowed; reviewer/auditor phases below must not see the style ref.)

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
• Freeze a Claim–Evidence–Figure Storyboard — each primary figure/table has a question, evidence locator, bounded conclusion, excluded conclusion, and unresolved analysis or user decision
• Design section structure (5-8 sections depending on paper type)
• Plan figure/table placement in narrative evidence order with data sources
• Scaffold citation structure
• GPT-5.4 reviews the plan for completeness

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

If — style-ref: <source> was passed in $ARGUMENTS and the helper succeeded above, append — style-ref: <source> to every writer-side sub-skill invocation in this pipeline (Phases 1, 2b, 3, 5). Do not append it to reviewer/auditor invocations (Phases 4.5, 4.7, 5.5, 5.8).

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
• GPT-5.4 reviews figure quality and captions

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

Scope: paper-figure covers data plots and comparison tables. Architecture diagrams, pipeline figures, and method illustrations are handled in Phase 2b below.

Phase 2b: Architecture & Illustration Generation

Skip this step entirely if illustration: false.

If the paper plan includes architecture diagrams, pipeline figures, audit cascades, or method illustrations, invoke the appropriate generator based on the illustration parameter:

When illustration: figurespec (default) — invoke /figure-spec:

/figure-spec "[architecture/workflow description from PAPER_PLAN.md]"

• Deterministic JSON → SVG vector rendering (editable, reproducible)
• Best for: system architecture, workflow pipelines, audit cascades, layered topology
• Output: figures/*.svg + figures/*.pdf (via rsvg-convert) + figures/specs/*.json
• No external API, runs fully local

If — style-ref: <source> was passed and the helper succeeded above, append — style-ref: <source> to the invocation below as well.

When illustration: gemini — invoke /paper-illustration:

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

• Claude plans → Gemini optimizes → Nano Banana Pro renders → Claude reviews (score ≥ 9)
• Best for: qualitative method illustrations, natural-style diagrams, result grids
• Output: figures/ai_generated/*.png
• Requires GEMINI_API_KEY environment variable

When illustration: mermaid — invoke /mermaid-diagram:

/mermaid-diagram "[method description from PAPER_PLAN.md]"

• Generates Mermaid syntax diagrams (flowchart, sequence, class, state, etc.)
• Best for: lightweight flowcharts, state machines, simple sequence diagrams
• Output: figures/*.mmd + figures/*.png
• Free, no API key needed

When illustration: codex-image2 — invoke /paper-illustration-image2:

/paper-illustration-image2 "[method description from PAPER_PLAN.md or NARRATIVE_REPORT.md]"

• Claude plans → Codex native image generation renders → Claude reviews (same multi-stage workflow as gemini, different renderer)
• Best for: users who want a GPT-image-style renderer without needing GEMINI_API_KEY; uses your existing Codex / ChatGPT Plus/Pro quota
• Output: figures/ai_generated/figure_final.png + latex_include.tex + review_log.json (emitted via the /paper-illustration-image2 SKILL's finalize step, which delegates to the canonical paper_illustration_image2.py helper resolved per integration-contract §2)
• Prerequisites (beyond debuffer's standard Claude Code + Codex coexistence): the local Codex app-server must be signed in (codex debug app-server send-message-v2 "ping" succeeds), and the dedicated MCP bridge must be registered once with claude mcp add codex-image2 -s user -- python3 ~/.claude/mcp-servers/codex-image2/server.py after copying mcp-servers/codex-image2/server.py there. Delegate the preflight to /paper-illustration-image2 (which resolves the helper via the canonical chain), or invoke the helper directly via the shim at tools/paper_illustration_image2.py preflight --workspace . to confirm before relying on this path.
• Experimental: this renderer shells through the Codex debug app-server, which Codex documents as an unstable surface. Prefer figurespec or gemini for production submission flows until codex-image2 stabilizes.

When illustration: false — skip entirely. All non-data figures must be created manually (draw.io, Figma, TikZ) and placed in figures/ before Phase 3.

Choosing the right mode:

• Formal architecture / workflow / topology figures → figurespec (default)
• Method concept illustrations with natural style, have GEMINI_API_KEY → gemini
• Method concept illustrations, prefer ChatGPT Plus/Pro quota over Gemini key → codex-image2
• Quick flowchart / state machine → mermaid
• Full manual control → false

These are complementary, not mutually exclusive: you can run multiple generators for different figures in the same paper by re-invoking with different illustration overrides.

Checkpoint: List generated vs manual figures.

📊 Figures complete:
- Data plots (auto, Phase 2): [list]
- Architecture/illustrations (auto, Phase 2b, mode=<illustration>): [list]
- 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?

Phase 3: LaTeX Writing

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

/paper-write "PAPER_PLAN.md"

If — style-ref: <source> was passed in $ARGUMENTS and the helper succeeded above, append — style-ref: <source> to the invocation: /paper-write "PAPER_PLAN.md — style-ref: <source>".

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"...)
• GPT-5.4 reviews each section for quality

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 4.3: Figure/Table Audit

Run /figure-table-audit "paper/" to check caption quality, reference order, visual readability, vector-format hygiene, and table presentation before the improvement loop.

Phase 4.4: Experiment Writeup Audit

Run /experiment-writeup-audit "paper/" to judge whether the experiment section is missing claim-critical studies or mainly needs stronger writeup.

Phase 4.5: Proof Verification (theory papers only)

Skip this phase if the paper contains no theorems, lemmas, or proofs.

if paper contains \begin{theorem} or \begin{lemma} or \begin{proof}:
    Run /proof-checker "paper/"
    This invokes GPT-5.4 xhigh to:
    - Verify all proof steps (hypothesis discharge, interchange justification, etc.)
    - Check for logic gaps, quantifier errors, missing domination conditions
    - Attempt counterexamples on key lemmas
    - Generate PROOF_AUDIT.md with issue list + severity

    If FATAL or CRITICAL issues found:
        Fix before proceeding to improvement loop
    If only MAJOR/MINOR:
        Proceed, improvement loop may address remaining issues
else:
    skip — no proofs, no action

Phase 4.6: Math Consistency Audit

Run /paper-math-consistency-audit "paper/" after the first successful compile and before the improvement loop whenever the manuscript has formulas, project-specific notation, math_commands.tex, equation labels, theorem-like statements, or appendix derivations.

if paper contains math macros, equation environments, labels/refs, or theorem-like blocks:
    Run /paper-math-consistency-audit "paper/"
    Check symbol definitions, macro conflicts, equation labels, notation drift,
    appendix restatements, and long-conversation formula consistency.

    If FAIL:
        Fix CRITICAL notation/formula conflicts before improvement loop
    If WARN:
        Queue findings into the improvement loop
else:
    skip in draft mode; under assurance=submission the audit emits NOT_APPLICABLE

Phase 4.7: Paper Claim Audit

Skip if no result files exist (e.g., survey/position papers with no experiments).

if results/*.json or results/*.csv or outputs/*.json exist:
    Run /paper-claim-audit "paper/"
    Fresh zero-context reviewer compares every number in the paper
    against raw result files. Catches rounding inflation, best-seed
    cherry-pick, config mismatch, delta errors.

    If FAIL:
        Fix mismatched numbers before improvement loop
    If WARN:
        Proceed, but flag for manual verification
else:
    skip — no experimental results to verify

Phase 5: Auto Improvement Loop

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

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

If — style-ref: <source> was passed in $ARGUMENTS and the helper succeeded above, append — style-ref: <source> to the invocation: /auto-paper-improvement-loop "paper/ — style-ref: <source>". The improvement loop's reviewer sub-agent will still NOT see the style ref (the loop's own SKILL forbids it); only the fix-implementation phase consumes it.

What this does (2 rounds):

Round 1: GPT-5.4 xhigh reviews the full paper → identifies CRITICAL/MAJOR/MINOR issues → Claude Code implements fixes → recompile → save main_round1.pdf

Round 2: GPT-5.4 xhigh re-reviews with conversation context → identifies remaining issues → Claude Code implements fixes → recompile → save main_round2.pdf

Typical improvements:

• Fix assumption-model mismatches
• Soften overclaims to match evidence
• Add missing interpretations and notation
• Strengthen limitations section
• 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. Location-aware thresholds: any main-body overfull blocks completion regardless of size; appendix overfulls block only if >10pt; bibliography overfulls block only if >20pt.

Phase 5.5: Final Paper Claim Audit (MANDATORY submission gate)

After /auto-paper-improvement-loop finishes, rerun /paper-claim-audit before the final report whenever the paper contains numeric claims and machine-readable raw result files exist.

Use the same detectors as Phase 4.7:

• numeric-claim regex over paper/main.tex and paper/sections/*.tex
• raw-evidence file search in results/, outputs/, experiments/, and figures/ for .json, .jsonl, .csv, .tsv, .yaml, or .yml

This phase is mandatory if both detectors are positive. It blocks the final report. If numeric claims exist but no raw result files are found, stop and warn the user before declaring the paper complete. If no numeric claims exist, skip.

NUMERIC_CLAIMS=$(rg -n -e '[0-9]+(\.[0-9]+)?\s*(%|\\%|±|\\pm|x|×)' \
  -e '(accuracy|BLEU|F1|AUC|mAP|top-1|top-5|error|loss|perplexity|speedup|improvement)' \
  paper/main.tex paper/sections 2>/dev/null || true)

RAW_RESULT_FILES=$(find results outputs experiments figures -type f \
  \( -name '*.json' -o -name '*.jsonl' -o -name '*.csv' -o -name '*.tsv' -o -name '*.yaml' -o -name '*.yml' \) 2>/dev/null | head -200)

if [ -n "$NUMERIC_CLAIMS" ] && [ -n "$RAW_RESULT_FILES" ]; then
    Run /paper-claim-audit "paper/"
    If FAIL:
        Fix mismatched numbers before the final report
elif [ -n "$NUMERIC_CLAIMS" ]; then
    Stop and warn: the paper contains numeric claims but no raw evidence files were found
fi

Empirical motivation: in a real submission run, the final paper claimed a narrower experiment grid than the raw JSON actually contained, and a tolerance value was rounded down past the actual relative error. Both were caught only after manual paper-claim-audit invocation in the final round; the improvement loop did not detect them.

Phase 5.6: Kill-Argument Adversarial Review (theory / scope-heavy papers)

After Phase 5.5 (claim audit) passes, run /kill-argument whenever the paper is theory-heavy or makes explicit scope/generality claims in the title or abstract. This is a final adversarial check that complements the claim/citation/proof audits: those verify local correctness (numbers match, cites resolve, theorems prove); kill-argument tests headline-level survival — whether the paper as a whole answers the worst rejection paragraph a senior area chair would write.

THEORY_ENV_COUNT=$(rg -c '\\begin\{(theorem|lemma|proposition|corollary)\}' paper/main.tex paper/sections 2>/dev/null | awk -F: '{s+=$2} END {print s+0}')
SCOPE_HINT=$(rg -i 'general(ization)?|broad|universal|across|any [A-Za-z]+ model|holds for' paper/sections/0*abstract* 2>/dev/null | head -1)

if [ "$THEORY_ENV_COUNT" -ge 5 ] || [ -n "$SCOPE_HINT" ]; then
    /kill-argument "paper/"
    KILL_VERDICT=$(jq -r '.verdict' paper/KILL_ARGUMENT.json)
    KILL_REASON=$(jq -r '.reason_code' paper/KILL_ARGUMENT.json)
fi

Gating (depends on the resolved assurance level from Phase 0):

Assurance level	THEORY/SCOPE detected	Behavior
submission	yes	MANDATORY. FAIL blocks the final report; WARN requires explicit user acknowledgment; BLOCKED/ERROR blocks the final report (cannot ship without an adversarial pass).
submission	no	Skip — KILL_ARGUMENT.json is written with verdict: NOT_APPLICABLE, reason_code: not_theory_or_scope_paper so the submission verifier sees a record.
internal / draft	yes	Advisory. Run if the user passed — kill-argument: true; otherwise skip. WARN/FAIL is logged but does not block.
internal / draft	no	Skip.

/kill-argument itself never edits the paper; it writes KILL_ARGUMENT.{md,json}. If still_unresolved critical points are surfaced, queue them for the next /auto-paper-improvement-loop round (Step 5.5 of that skill auto-merges the findings into its fix list).

Why this is the right place: Phase 5 (loop) optimizes for score, Phase 5.5 (claim audit) verifies numbers, Phase 5.8 (citation audit) verifies cites — none of these catches the case where every local component is correct but the paper still oversells what it actually proves. Kill-argument is the dedicated headline-scope check.

Phase 5.7: Final Math Consistency Audit (submission gate)

After the final claim/scope fixes and before /citation-audit, rerun /paper-math-consistency-audit "paper/" whenever the paper contains formulas, labels, notation macros, or theorem-like blocks. This catches drift introduced by late edits, especially across math_commands.tex, Content/ or sections/, and appendix files.

Under assurance=submission, this audit must always emit paper/MATH_CONSISTENCY_AUDIT.json; detector-negative papers emit NOT_APPLICABLE. FAIL, BLOCKED, or ERROR blocks the Final Report.

Phase 5.8: Citation Audit (submission gate)

After the final paper-claim-audit passes, run /citation-audit to verify every \cite{...} along three axes: existence, metadata correctness, and context appropriateness. This is the fourth and final layer of the evidence-and-claim assurance stack (experiment-audit → result-to-claim → paper-claim-audit → citation-audit).

if paper/references.bib (or paper.bib) exists and contains entries cited from sec/*.tex:
    Run /citation-audit "paper/"
    Fresh cross-family reviewer (gpt-5.5 via Codex MCP) with web/DBLP/arXiv lookup
    verifies each entry:
      (i)   EXISTENCE — paper resolves at claimed arXiv ID / DOI / venue
      (ii)  METADATA — author names, year, venue, title match canonical sources
      (iii) CONTEXT — cited paper actually establishes the claim it supports

    Output:
      - CITATION_AUDIT.md (human-readable per-entry verdict report)
      - CITATION_AUDIT.json (machine-readable verdict ledger)
      - Per-entry verdicts: KEEP / FIX / REPLACE / REMOVE

    If any REPLACE or REMOVE verdicts:
        Surface to user for human approval — never auto-modify content claims
    If only FIX verdicts (metadata corrections):
        Apply with user confirmation, then recompile
    If all KEEP:
        Pass — bibliography clean for submission
else:
    skip — no bib file or no citations

Why this is the most diagnostic of the four audit layers: wildly fake citations are easy to spot. The dangerous failure mode is a real paper used to support a claim it does not actually establish (wrong-context citations) — these slip past metadata-only checks and damage submission credibility. Run cost is wall-clock heavy (web lookup per entry); run once per submission, not per save.

Empirical motivation: in a real submission run, several real papers were cited in contexts they did not actually support, and at least one bib entry shipped with author = "Anonymous" because the metadata had not been resolved. None were caught by the improvement loop or numeric claim audit; only fresh web-lookup review surfaced them.

Phase 6: Final Report

Phase 6.0 — Submission Gate

Before writing the Final Report, resolve the active assurance level. This uses the same derivation rule as Phase 0 so a run where Phase 0 was skipped or its write failed cannot silently downgrade a beast / max / — assurance: submission invocation back to draft.

Resolution at the gate (re-derive; do not trust .debuffer_skills/assurance.txt alone):

1. Parse $ARGUMENTS for an explicit — assurance: draft | submission or an — effort: lite | balanced | max | beast directive.
2. Derive the expected level:
   • explicit assurance: wins
   • else lite / balanced → draft, max / beast → submission
   • else draft
3. Read paper/.debuffer_skills/assurance.txt. If the file is missing, write it now with the derived level.
4. If the file's value disagrees with the derived level (e.g. file says draft but $ARGUMENTS says beast), overwrite the file with the derived level and surface a one-line warning in-chat: ⚠️ .debuffer_skills/assurance.txt was draft but $ARGUMENTS says submission; overriding.
5. Use the re-derived level as authoritative for the rest of Phase 6.

# Final authoritative value, written and read from the same source
ASSURANCE=<derived-from-$ARGUMENTS>        # draft | submission
mkdir -p paper/.debuffer_skills
echo "$ASSURANCE" > paper/.debuffer_skills/assurance.txt

If ASSURANCE=draft, skip directly to the Final Report template below — current behavior, no change for the default balanced user.

If ASSURANCE=submission, run the pre-flight checklist below, then the verifier. The verifier's exit code is the source of truth — do NOT self-declare "audits complete" based on conversation memory.

Submission pre-flight checklist

Print this checklist verbatim at the start of Phase 6.0 and confirm each row before proceeding. This resists the common failure mode of the model skipping audits while claiming to have run them.

📋 Submission audits required before Final Report:
   [ ] 1. /proof-checker        → paper/PROOF_AUDIT.json
   [ ] 2. /paper-math-consistency-audit → paper/MATH_CONSISTENCY_AUDIT.json
   [ ] 3. /paper-claim-audit    → paper/PAPER_CLAIM_AUDIT.json
   [ ] 4. /kill-argument        → paper/KILL_ARGUMENT.json
   [ ] 5. /citation-audit       → paper/CITATION_AUDIT.json
   [ ] 6. Resolve $AUDIT_VERIFIER per integration-contract.md §2 (Policy A),
          then: bash "$AUDIT_VERIFIER" paper/ --assurance submission
   [ ] 7. Block Final Report iff verifier exit code != 0

The resolver in "Running the verifier" below tries .debuffer_skills/tools/verify_paper_audits.sh (created by install_aris.sh), then tools/verify_paper_audits.sh (in-repo run), then $ARIS_REPO/tools/verify_paper_audits.sh (env-var-set path). The chain always tries layers 1 → 2 → 3 in order; setting export ARIS_REPO=~/… only ensures layer 3 has a valid target if layers 1 and 2 are absent.

Invoking the mandatory audits

Each sub-audit runs in a fresh Codex thread (never codex-reply, never pass prior audit output as context — this preserves reviewer independence per shared-references/reviewer-independence.md).

Each sub-audit always emits its JSON artifact, even when the content detector is negative. A detector-negative run emits verdict NOT_APPLICABLE; a silent skip is forbidden. See the "Submission artifact emission" section of each audit's SKILL.md.

Order:

1. /proof-checker "paper/" → writes paper/PROOF_AUDIT.json (emits NOT_APPLICABLE if the paper contains no theorems / lemmas / proofs)
2. /paper-math-consistency-audit "paper/" → writes paper/MATH_CONSISTENCY_AUDIT.json (emits NOT_APPLICABLE if the paper contains no math macros, equations, labels, or theorem-like blocks)
3. /paper-claim-audit "paper/" → writes paper/PAPER_CLAIM_AUDIT.json (emits NOT_APPLICABLE if the paper has no numeric claims; emits BLOCKED if numeric claims exist but raw result files are missing)
4. /kill-argument "paper/" → writes paper/KILL_ARGUMENT.json (emits NOT_APPLICABLE if the paper is not theory/scope-heavy)
5. /citation-audit "paper/" → writes paper/CITATION_AUDIT.json (emits NOT_APPLICABLE if no .bib file or no \cite{...} usage)

Running the verifier

Resolve $AUDIT_VERIFIER via the canonical strict-safe chain (see shared-references/integration-contract.md §2, Policy A — gate). Under assurance: submission the verifier is load-bearing: if the helper is unresolved the SKILL aborts the Final Report rather than producing an unverified submission-ready claim.

# Resolve the audit verifier (Policy A — gate).
cd "$(git rev-parse --show-toplevel 2>/dev/null || pwd)" || exit 1
if [ -z "${ARIS_REPO:-}" ] && [ -f .debuffer_skills/installed-skills.txt ]; then
    ARIS_REPO=$(awk -F'\t' '$1=="repo_root"{print $2; exit}' .debuffer_skills/installed-skills.txt 2>/dev/null) || true
fi
AUDIT_VERIFIER=".debuffer_skills/tools/verify_paper_audits.sh"
[ -f "$AUDIT_VERIFIER" ] || AUDIT_VERIFIER="tools/verify_paper_audits.sh"
[ -f "$AUDIT_VERIFIER" ] || { [ -n "${ARIS_REPO:-}" ] && AUDIT_VERIFIER="$ARIS_REPO/tools/verify_paper_audits.sh"; }
[ -f "$AUDIT_VERIFIER" ] || {
  echo "ERROR: verify_paper_audits.sh not resolved at .debuffer_skills/tools/, tools/, or \$ARIS_REPO/tools/." >&2
  echo "       assurance=submission requires the verifier; aborting Final Report." >&2
  echo "       Fix: rerun bash tools/install_aris.sh, export ARIS_REPO, or copy the helper to tools/." >&2
  exit 1
}

bash "$AUDIT_VERIFIER" paper/ --assurance submission

• Exit 0 — All mandatory audits present, JSON schema-valid, hashes fresh, no blocking verdicts. Proceed to the Final Report below.
• Exit 1 — Surface paper/.debuffer_skills/audit-verifier-report.json to the user verbatim, refuse to generate the Final Report, and list the specific remediation for each failing row:
  • MISSING → rerun that audit
  • STALE → paper files edited after the audit ran; rerun the affected audit
  • BLOCKING_VERDICT (FAIL / BLOCKED / ERROR) → fix the underlying issue, then rerun the audit
  • SCHEMA_INVALID → audit artifact malformed; rerun the audit

The verifier is cheap to rerun (< 1 s). After fixing any issue, rerun it before claiming green.

Optional hardening (not default)

Teams that want hook-level enforcement — i.e., the harness physically prevents a Stop event while the verifier is red — can register a Stop hook in ~/.claude/settings.json:

{
  "hooks": {
    "Stop": [
      {"command": "bash <ARIS_REPO>/tools/verify_paper_audits.sh paper/ --assurance submission"}
    ]
  }
}

This is documented here, not required. Phase 6.0's verifier-as-truth pattern is the default repo behavior.

---

Phase 6.1 — Final Report (runs only after the submission gate is green, or directly if assurance=draft)

# Paper Writing Pipeline Report

**Input**: [NARRATIVE_REPORT.md or topic]
**Venue**: [ICLR/NeurIPS/ICML/CVPR/ACL/AAAI/ACM/IEEE_JOURNAL/IEEE_CONF]
**Assurance**: [draft | submission]
**Submission-ready**: [yes | no]   <!-- yes iff assurance=submission AND verifier exit 0 -->
**Date**: [today]

## Pipeline Summary

| Phase | Status | Output |
|-------|--------|--------|
| 0. Assurance Setup | ✅ | paper/.debuffer_skills/assurance.txt = [draft\|submission] |
| 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 |
| 4.5 Proof Audit | [PASS\|WARN\|FAIL\|NOT_APPLICABLE\|BLOCKED\|ERROR] | PROOF_AUDIT.{md,json} |
| 5.7 Math Consistency Audit | [PASS\|WARN\|FAIL\|NOT_APPLICABLE\|BLOCKED\|ERROR] | MATH_CONSISTENCY_AUDIT.{md,json} |
| 5.5 Paper Claim Audit | [PASS\|WARN\|FAIL\|NOT_APPLICABLE\|BLOCKED\|ERROR] | PAPER_CLAIM_AUDIT.{md,json} |
| 5.8 Citation Audit | [PASS\|WARN\|FAIL\|NOT_APPLICABLE\|BLOCKED\|ERROR] | CITATION_AUDIT.{md,json} |
| 6.0 Assurance Verifier | [OK\|STALE\|BLOCKING_VERDICT\|HAS_ISSUES\|SCHEMA_INVALID\|MISSING] per audit; exit [0\|1] overall (N/A if draft) | .debuffer_skills/audit-verifier-report.json |

## 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
- paper/PROOF_AUDIT.{md,json} — Proof-obligation verification (always emitted at `assurance=submission`; `NOT_APPLICABLE` when no theorems)
- paper/MATH_CONSISTENCY_AUDIT.{md,json} — Formula, notation, macro, label, and appendix restatement consistency (always emitted at `assurance=submission`; `NOT_APPLICABLE` when no math-like content)
- paper/PAPER_CLAIM_AUDIT.{md,json} — Numerical claim verification (always emitted at `assurance=submission`; `NOT_APPLICABLE` when no numeric claims; omitted in `draft` mode if Phase 5.5 detector was negative)
- paper/CITATION_AUDIT.{md,json} — Bibliography verification (always emitted at `assurance=submission`; `NOT_APPLICABLE` when no `.bib` or no `\cite{...}`; omitted in `draft` mode if Phase 5.8 detector was negative)
- paper/.debuffer_skills/audit-verifier-report.json — External verifier report (submission only)

## 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 docs/project/OUTPUT_MANIFEST.md
• Output Language Protocol — note: paper-writing always outputs English LaTeX for venue submission

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"                  -> find/refine ideas
/research-blueprint                          -> freeze theory, method, and experiment gates
/autodl-hpc or /experiment-plan              -> formal runs, not local heavy compute
/experiment-audit or evidence ledger update  -> map formal results to claims
/paper-plan "docs/paper/NARRATIVE_REPORT.md" -> plan only after evidence gate passes
/paper-writing "docs/paper/PAPER_PLAN.md"    -> draft only after manuscript entry gate passes

If the project has only smoke, pilot, toy, or validation results, stay in the
experiment/audit phases and do not run this skill.

Typical Timeline

Phase	Duration	Can sleep?
1. Paper Plan	5-10 min	No
2. Figures	5-15 min	No
3. LaTeX Writing	15-30 min	Yes ✅
4. Compilation	2-5 min	No
5. Improvement	15-30 min	Yes ✅

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