paper-poster-html

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Paper Poster (HTML): measurement-gated poster generation

论文海报（HTML版）：基于测量校验的海报生成

One HTML file styled for an exact print canvas (

@page { size: W H }

), rendered to PDF via Playwright print emulation. Iterate by measuring, not eyeballing — the screen preview lies; only print emulation at the correct viewport tells the truth. Core gate machinery is adapted from posterly (MIT, © 2026 Ruishuo Chen — see

NOTICE.md

and

LICENSES/posterly-MIT.txt

); ARIS adds style discipline gates, figure-provenance gates, the cross-model review loop, and the anti-patch-loop fix vocabulary.

单个HTML文件，针对精确打印画布进行样式设置（

@page { size: W H }

NOTICE.md

和

LICENSES/posterly-MIT.txt

）；ARIS新增了样式规范校验、图表来源校验、跨模型评审循环以及防补丁循环的修复词汇表。

Why this skill exists (the failure it prevents)

该技能存在的原因（解决的痛点）

A predecessor pipeline produced a poster with 30+ colors, zero real paper figures, a screen-pixel canvas, and tiny formulas floating in oversized boxes, then spent 12+ review rounds making it worse — each round added a new badge color or bespoke SVG patch. The cure is structural, not exhortative:

Hard gates run before any aesthetic opinion (alignment, style, assets must PASS first — a reviewer never sees an unmeasured poster).
A closed fix vocabulary — visual-review fixes can only touch design tokens, whole catalogued components, content rebalance, assets, or canvas choice. New inline styles / new hex values / bespoke decorations are structurally forbidden.
Two-hue discipline as a machine check, not a style suggestion.
Real paper figures with provenance manifest, or the gate fails.

此前的流水线生成的海报存在30余种颜色、无真实论文图表、基于屏幕像素画布、公式在超大框中悬浮等问题，后续经过12轮以上的评审后效果反而更差——每一轮都新增了新的徽章颜色或定制SVG补丁。解决方案是结构性的，而非劝诫性的：

严格校验优先于任何审美意见（对齐、样式、资源必须先通过校验——评审者永远不会看到未经过测量校验的海报）。
封闭的修复词汇表——视觉评审的修复只能修改设计令牌、已分类的完整组件、内容平衡、资源或画布选择。从结构上禁止新增内联样式/新十六进制颜色值/定制装饰。
双色调规范作为机器校验项，而非风格建议。
带来源清单的真实论文图表，否则校验不通过。

Mental model

心智模型

paper (.tex / PDF) ──► content plan + claim→evidence audit (codex, fresh)
                              │
   figures extracted ─────────┤  FIGURE_MANIFEST.json (provenance, sha256)
   (real paper figures ONLY)  ▼
   template scaffold ──► fill ──► run_gates.py            ◄─── HARD, loop here
                                  preflight → style → asset → measure → polish
                              │ all hard gates PASS
                              ▼
                    Claude visual review (≤3 issues × ≤3 rounds, fix-vocabulary only)
                              │ score ≥ 9
                              ▼
                    codex final cross-model review (fresh thread, full HTML+PDF)
                              │ pass
                              ▼
                    verify-final → poster.pdf + GATE_REPORT.json

paper (.tex / PDF) ──► 内容规划 + 论点→证据审核（codex，全新线程）
                              │
   提取的图表 ─────────┤  FIGURE_MANIFEST.json（来源信息，sha256）
   （仅真实论文图表）  ▼
   模板脚手架 ──► 填充内容 ──► run_gates.py            ◄─── 严格校验，在此循环
                                  预检 → 样式 → 资源 → 测量 → 优化
                              │ 所有严格校验通过
                              ▼
                    Claude视觉评审（≤3个问题 × ≤3轮，仅使用修复词汇表）
                              │ 评分≥9
                              ▼
                    codex最终跨模型评审（全新线程，完整HTML+PDF）
                              │ 通过
                              ▼
                    最终验证 → poster.pdf + GATE_REPORT.json

Constants

常量定义

SKILL_SCRIPTS =
```
${CLAUDE_SKILL_DIR}/scripts
```
— all helpers are single-owner and ship inside this skill (Arch C). If the directory is missing the install is broken: abort and tell the user to re-install the skill (Policy A — the gates ARE the skill; never improvise replacements).
REVIEWER_MODEL =
```
gpt-5.6-sol
```
, reasoning
```
xhigh
```
, fresh thread per review call (
```
mcp__codex__codex
```
, never
```
codex-reply
```
across review boundaries).
CANVAS — from the venue's official spec, looked up live in Phase 0. Never assume. (Known anchor: ICLR 2026 main = 185×90 cm landscape per its official printing service; ICML/NeurIPS commonly 60×36 in landscape; workshop posters often 61×91 cm portrait. Specs change yearly — verify.)
PALETTE — default =
```
templates/tokens/generic.json
```
(slate-blue
```
#2D5F8B
```
accent
- gold
```
#C9A24A
```
  highlight + neutrals) for all venues. Venue packs are opt-in via
```
— venue-colors: true
```
  . Purple-dominant accents (hue 250–285) are banned unless the user passes
```
— allow-purple: true
```
  .
AUTO_PROCEED = false — wait for explicit confirmation at every 🚦 checkpoint.
OUTPUT_DIR =
```
poster_html/
```
in the working directory.

SKILL_SCRIPTS =
```
${CLAUDE_SKILL_DIR}/scripts
```
—— 所有辅助脚本均为单一所有者维护，并随该技能一同发布（架构C）。若目录缺失则安装失败：终止流程并告知用户重新安装该技能（策略A——校验机制是技能的核心；切勿临时替代）。
REVIEWER_MODEL =
```
gpt-5.6-sol
```
，推理等级
```
xhigh
```
，每次评审调用使用全新线程（
```
mcp__codex__codex
```
，评审边界间绝不能使用
```
codex-reply
```
）。
CANVAS —— 来自会议官方规范，在第0阶段实时查询。切勿假设已知。（参考示例：ICLR 2026主会场海报为185×90厘米横版，来自其官方印刷服务；ICML/NeurIPS通常为60×36英寸横版； workshop海报常为61×91厘米竖版。规范每年变化——务必验证。）
PALETTE —— 默认值 =
```
templates/tokens/generic.json
```
（石板蓝
```
#2D5F8B
```
强调色 + 金色
```
#C9A24A
```
高亮色 + 中性色），适用于所有会议。可通过
```
— venue-colors: true
```
选择启用会议专属配色包。除非用户传入
```
— allow-purple: true
```
，否则禁止使用紫色系强调色（色相250–285）。
AUTO_PROCEED = false —— 在每个🚦检查点等待用户明确确认。
OUTPUT_DIR = 工作目录下的
```
poster_html/
```
。

Workflow

工作流程

Phase 0 — Resume, dependencies, venue spec

第0阶段 —— 恢复、依赖项、会议规范

Resume: if
```
poster_html/POSTER_STATE.json
```
exists with
```
status: in_progress
```
(< 24 h), resume from the saved phase.
Dependencies (degradation chain, in order):
- Playwright + bundled Chromium → if missing,
```
python3 -m playwright install chromium
```
  → if install fails but system Chrome exists, scripts fall back to
```
channel="chrome"
```
  → if all fail: you may produce the content plan and scaffold only, label everything "not print verified", and must NOT emit a final PDF.
- ```
pdfinfo
```
  missing → PyMuPDF reads PDF dimensions. At least one of pdftoppm / PyMuPDF must exist for PNG review renders.
- MathJax: download
```
tex-svg.js
```
  once into
```
poster_html/assets/mathjax/
```
  and reference it locally in the HTML. CDN is acceptable only for drafts; the measure gate hard-fails on unrendered MathJax either way.
Venue spec lookup (live): consult the venue's official poster-instructions page (search + fetch). Extract dimensions, orientation, font floor, logo policy, anonymity rules, file format. Record
```
{spec, source_url, retrieved}
```
into
```
POSTER_STATE.json
```
— specs change yearly; never reuse a cached spec silently.

🚦 Checkpoint: echo the venue spec table (canvas, orientation, source URL) and the chosen template. Wait.

恢复流程：若
```
poster_html/POSTER_STATE.json
```
存在且
```
status: in_progress
```
（创建时间<24小时），则从保存的阶段恢复。
依赖项（降级顺序）：
- Playwright + 捆绑版Chromium → 若缺失，执行
```
python3 -m playwright install chromium
```
  → 若安装失败但系统存在Chrome，脚本将回退至
```
channel="chrome"
```
  → 若全部失败：仅可生成内容规划和脚手架，标记所有内容为**“未经过打印验证”**，且不得生成最终PDF。
- ```
pdfinfo
```
  缺失 → 使用PyMuPDF读取PDF尺寸。必须至少存在pdftoppm / PyMuPDF中的一个，用于生成PNG评审预览图。
- MathJax：将
```
tex-svg.js
```
  下载一次至
```
poster_html/assets/mathjax/
```
  ，并在HTML中本地引用。仅草稿可使用CDN；无论如何，测量校验会因未渲染的MathJax而失败。
实时查询会议规范：查阅会议官方海报指南页面（搜索+获取）。提取尺寸、方向、最小字体要求、Logo政策、匿名规则、文件格式。将
```
{spec, source_url, retrieved}
```
记录至
```
POSTER_STATE.json
```
——规范每年变化；切勿静默复用缓存的规范。

🚦 检查点：输出会议规范表格（画布、方向、来源URL）和选定的模板。等待用户确认。

Phase 0.5 — Design discovery (one AskUserQuestion batch)

第0.5阶段 —— 设计确认（一次用户提问批量完成）

Ask once, ≤4 questions: layout template (from

templates/README.md

), palette (default generic pack / venue pack / custom within constraints), logos + venue mark (paths or "none" — never fabricate; check the venue's logo policy), QR target (paper / code / project page / none — generate offline with

qrencode

or python-

qrcode

; never a remote QR-service URL). Persist answers in

POSTER_STATE.json

design_decisions

— re-read before any later "improvement" so deliberate choices are never reverted.

仅提问一次，最多4个问题：布局模板（来自

templates/README.md

）、配色方案（默认通用包/会议专属包/符合约束的自定义方案）、Logo+会议标识（路径或“无”——切勿伪造；需符合会议Logo政策）、二维码目标（论文/代码/项目页面/无——使用

qrencode

或python-

qrcode

离线生成；禁止使用远程二维码服务URL）。将答案保存至

POSTER_STATE.json

的

design_decisions

字段——后续进行任何“优化”前需重新读取，避免回退用户的明确选择。

Phase 1 — Paper ingest, content plan, claim audit

第1阶段 —— 论文导入、内容规划、论点审核

Read the paper source (
```
.tex
```
ideal; PDF otherwise). Extract: title/authors/affils, the 3–5 headline numbers, core method (equations verbatim), main results (tables/figures and what they show), takeaways. Build
```
poster_html/POSTER_CONTENT_PLAN.md
```
— what goes in which column, word budget per card. Target density (excluding table cells, captions, author line, footer): standard poster 550–850 words; dense theory+empirical poster 750–1050 words, allowed only when ≥2 compact components are used (
```
eqn-anatomy
```
,
```
flow-strip
```
,
```
derived-col
```
,
```
claim-pills
```
,
```
keybox--4
```
). Warn yourself below 500 words on a 4-column landscape (it will read as sparse next to professionally dense posters) unless the template is hero/visual-first; warn above 1100 unless the user asked for dense mode. Bullets ≤ 8 words when possible — density comes from structure, not long prose. Prefer compact structure over prose: if the paper contains an explicit objective, algorithm, theorem mechanism, or baseline comparison, extract at least two of: (1) empirical objective / loss stack; (2) term-by-term equation anatomy; (3) a method-flow strip grounded in paper variables; (4) a derived-Δ column for method-vs-baseline rows; (5) a 4-up implementation/theory keybox; (6) a claim/evidence pill table for numeric-heavy posters. Do not invent an algorithm. If the paper has only an objective, label the component "objective flow" or "loss anatomy", never "algorithm".
Cross-model content audit (fresh codex thread,
```
xhigh
```
): give it the content plan path + paper source path(s) — paths only, no summaries — and ask for a claim→evidence table:
```
| claim on poster | paper file:line | paper says (verbatim) | match? |
```
with match ∈ {OK, NUMERIC-MISMATCH, OVERCLAIM, MISSING-PRECONDITION, NOT-IN-PAPER, SCOPE-NARROWED}. Save to
```
poster_html/CLAIM_EVIDENCE.md
```
.
Fix every non-OK row or record it as a user-acknowledged tradeoff.

🚦 Checkpoint: content plan + audit summary. Wait.

读取论文源文件（优先
```
.tex
```
；其次PDF）。提取：标题/作者/机构、3–5个核心数据、核心方法（公式原文）、主要结果（表格/图表及其说明）、结论。生成
```
poster_html/POSTER_CONTENT_PLAN.md
```
——内容分布到哪些列，每个卡片的字数预算。目标密度（不含表格单元格、标题、作者栏、页脚）：标准海报550–850词；密集型理论+实证海报750–1050词，仅当使用≥2个紧凑组件时允许（
```
eqn-anatomy
```
,
```
flow-strip
```
,
```
derived-col
```
,
```
claim-pills
```
,
```
keybox--4
```
）。若4列横版海报字数低于500词（除非模板为视觉优先型），需发出警告；若超过1100词且未启用密集模式，也需警告。尽可能使用≤8词的项目符号——密度来自结构，而非冗长文本。优先选择紧凑结构而非散文：若论文包含明确的目标、算法、定理机制或基线对比，至少提取以下两项：(1) 实证目标/损失栈；(2) 逐术语公式解析；(3) 基于论文变量的方法流程条；(4) 方法vs基线的差异对比列；(5) 4栏式实现/理论关键框；(6) 数值密集型海报的论点/证据胶囊表格。切勿发明算法。若论文仅包含目标，将组件标记为“目标流程”或“损失解析”，绝不能标记为“算法”。
跨模型内容审核（全新codex线程，
```
xhigh
```
推理等级）：传入内容规划路径+论文源文件路径（仅传路径，无需摘要），要求生成论点→证据表格：
```
| 海报论点 | 论文文件:行号 | 论文原文 | 匹配度 |
```
，匹配度可选值为{OK, NUMERIC-MISMATCH, OVERCLAIM, MISSING-PRECONDITION, NOT-IN-PAPER, SCOPE-NARROWED}。保存至
```
poster_html/CLAIM_EVIDENCE.md
```
。
修复所有非OK行，或记录为用户确认的权衡项。

🚦 检查点：输出内容规划+审核摘要。等待用户确认。

Phase 2 — Real paper figures (provenance-gated)

第2阶段 —— 真实论文图表（来源校验）

Source preference chain:

Paper source
```
figures/
```
(vector SVG/PDF → convert to SVG via
```
inkscape
```
/
```
pdf2svg
```
if available, else rasterize ≥ 2× rendered px).
PDF-only:
```
extract_pdf_figures.py contact-sheet
```
+
```
auto
```
to list candidate regions → pick crops (🚦 human confirms crop choices) →
```
crop
```
at 300–450 DPI.
Last resort: user supplies explicit
```
page,x0,y0,x1,y1
```
bboxes.

Then

preprocess_figures.py --autocrop

every asset. Every paper-derived image gets a

FIGURE_MANIFEST.json

entry (source hash, page, bbox, dpi, sha256, natural_px) and is embedded as

<img data-source="paper" data-asset-id="...">

Hard rule: ≥ 2 paper-derived visuals or the asset gate fails. Theory-only papers may waive the total-area rule (

--waive-total-area

) at a human checkpoint — never silently. Never draw bespoke decorative SVG "figures" as substitutes.

Figure-area bands (asset gate, fractions of body): total target 14–22 % (warn < 12 % / > 24 %, hard < 10 % / > 28 %); per ordinary figure target 4–8 % (warn

10 %, hard > 13 %);
figure--duo
combined 8–12 %. Hero templates pass
--hero
(centerpiece may take 30–40 %). The failure mode is symmetric: too small reads as decoration, too big crowds out content. Sibling figures that share axes or tell a before→after story belong in one
figure--duo
card, not two cards.

来源优先级：

论文源文件的
```
figures/
```
目录（矢量SVG/PDF → 若可用，通过
```
inkscape
```
/
```
pdf2svg
```
转换为SVG，否则以≥2倍渲染像素光栅化）。
仅PDF文件：执行
```
extract_pdf_figures.py contact-sheet
```
+
```
auto
```
列出候选区域 → 选择裁剪区域（🚦 人工确认裁剪选择） → 以300–450 DPI裁剪。
最后手段：用户提供明确的
```
page,x0,y0,x1,y1
```
边界框。

然后对所有资源执行

preprocess_figures.py --autocrop

。每个来自论文的图像都会在

FIGURE_MANIFEST.json

中生成条目（源文件哈希、页码、边界框、DPI、sha256、原始像素尺寸），并以

<img data-source="paper" data-asset-id="...">

的形式嵌入。

严格规则：至少包含2个来自论文的可视化内容，否则资源校验失败。纯理论论文可在人工检查点豁免总面积规则（

--waive-total-area

）——切勿静默豁免。切勿绘制定制装饰性SVG“图表”作为替代。

图表面积范围（资源校验，占主体区域的比例）：总目标14–22 %（低于12%/高于24%时警告，低于10%/高于28%时校验失败）；单张普通图表目标4–8 %（高于10%时警告，高于13%时校验失败）；

figure--duo

组合图表8–12 %。视觉优先模板可传入

--hero

（中心图表可占30–40%）。两种极端均为失败模式：过小会显得像装饰，过大会挤占内容空间。共享坐标轴或展示前后对比的图表应放入同一个

figure--duo

卡片，而非两个独立卡片。

Phase 3 — Scaffold + tokens

第3阶段 —— 脚手架 + 令牌

cp templates/<chosen>.html poster_html/poster.html

; retarget

@page

.poster

dims to the venue canvas (two edits, same values); apply the chosen token pack onto the

:root

DESIGN TOKENS block; fill content per the plan; embed manifest figures. Run

preflight

style_check

— both must PASS before any layout iteration. (A fresh scaffold is expected to fail

measure

— that gate judges a filled poster.)

执行

cp templates/<chosen>.html poster_html/poster.html

；将

@page

.poster

尺寸修改为会议画布尺寸（两处修改，值相同）；将选定的令牌包应用到

:root

设计令牌块；根据规划填充内容；嵌入带来源清单的图表。执行

preflight

style_check

——两者均需通过才能进行布局迭代。（全新脚手架预期会

measure

校验失败——该校验针对填充后的海报。）

Phase 4 — Layout hard loop

第4阶段 —— 布局严格循环

After every layout change:

bash

python3 "$SKILL_SCRIPTS/run_gates.py" poster_html/poster.html \
    --tokens <pack.json> --manifest poster_html/FIGURE_MANIFEST.json \
    --report poster_html/GATE_REPORT.json

Canonical order: preflight → style → asset → measure → polish. Targets: column-bottom spread < 5 px (aim < 3), footer gap ∈ [30, 50] px, intercard gap ∈ [12, 50] px, canvas-fill ∈ [95, 101] %, poster bbox aligned to page within ±2 px. Fix guidance for each failure mode lives in the gate output and

templates/COMPONENTS.md

. Do not proceed while any hard gate fails. Do not let a reviewer see an unmeasured poster. Balance under-filled columns with content from the paper (Gate C), never with whitespace,

space-between

, or stretched cards.

每次布局修改后：

bash

python3 "$SKILL_SCRIPTS/run_gates.py" poster_html/poster.html \
    --tokens <pack.json> --manifest poster_html/FIGURE_MANIFEST.json \
    --report poster_html/GATE_REPORT.json

标准顺序：预检 → 样式 → 资源 → 测量 → 优化。目标：列底部间距<5 px（目标<3 px），页脚间隙∈[30, 50] px，卡片间间隙∈[12, 50] px，画布填充率∈[95, 101] %，海报边界框与页面对齐误差±2 px内。每种失败模式的修复指南可在校验输出和

templates/COMPONENTS.md

中找到。任何严格校验未通过时不得继续。切勿让评审者看到未经过测量校验的海报。 使用论文中的内容填充列的空白（规则C），切勿使用空白、

space-between

或拉伸卡片。

Phase 5 — Claude visual review (gated aesthetics)

第5阶段 —— Claude视觉评审（审美校验）

Render and read the result yourself:

bash

python3 "$SKILL_SCRIPTS/render_preview.py" poster_html/poster.html
pdftoppm -r 100 poster_html/poster_preview.pdf poster_html/review_full -png -f 1 -l 1

自行渲染并查看结果：

bash

python3 "$SKILL_SCRIPTS/render_preview.py" poster_html/poster.html
pdftoppm -r 100 poster_html/poster_preview.pdf poster_html/review_full -png -f 1 -l 1

plus 2-4 region crops at higher res (header / one column / equations) via PIL

另外通过PIL生成2-4个高分辨率区域裁剪图（页眉/某一列/公式）


**Calibrate first** (`../shared-references/taste-calibration.md`): if
**human-curated** `references/good/` + `references/bad/` exist under this skill
dir (or the project supplies its own pair), score those 3+3 reference posters
on the axes below BEFORE the target, anchoring the scale. Never select, search
for, or generate anchors yourself; if no reference sets exist, proceed
uncalibrated and mark `CALIBRATION: none` — never fabricate anchor scores.
Axes (weights sum 1.0): Design 0.35 · Craft 0.30 · Functionality 0.20 ·
Originality 0.15. Mapping: `SCORE = min(round(1 + 9 × COMPOSITE), lowest
triggered cap)` — caps apply AFTER the mapping, and the loop's `Score ≥ 9`
threshold below always reads this final capped `SCORE`, never the raw
composite.

Score strictly 1–10. **Critical caps** (hard floors — a calibrated composite
never overrides them): < 2 real paper figures → ≤ 3; broken canvas /
clipped content / unreadable math → ≤ 4; ≥ 4 visible hue families or gradient-heavy
header → ≤ 4; large blank cards or columns → ≤ 5; fabricated visual claim → ≤ 3.
Checks: posterly-showcase gestalt (would this hang next to a professionally designed
poster without looking like a patched dashboard?), single-accent discipline, real
figures readable and central, print hierarchy (title → headline stats → figures →
detail), column fill, **equation prominence** (no tiny math in oversized boxes),
serif-body/sans-display pairing, no gradient kitsch, component consistency, 60-second
narrative. Output format:

SCORE: N/10 (= min(round(1 + 9 × COMPOSITE), lowest cap); drives the loop) COMPOSITE: 0.xx (weighted; list the four per-axis scores) CALIBRATION: anchored | none GAP: <which reference poster the target falls short of / exceeds, on which axis, and why — one paragraph; omit only when CALIBRATION: none> CAPS_TRIGGERED: ... TOP_ISSUES: (max 3) ALLOWED_FIX_TYPE per issue: token | component | rebalance | asset | template/canvas PATCH_LOOP_RISK: low | medium | high


Loop: fix (fix vocabulary below) → re-run Phase 4 gates → re-score. **≤ 3 issues per
round, ≤ 3 rounds.** Score ≥ 9 → Phase 6. Still < 9 after 3 rounds → STOP patching;
escalate to template / canvas / content re-choice (back to Phase 3) or a human
decision. Never enter round 4 of cosmetic patching.


**先校准**（`../shared-references/taste-calibration.md`）：若该技能目录下存在**人工筛选**的`references/good/` + `references/bad/`参考海报（或项目提供了自己的参考对），在评估目标海报前，先对这些3+3个参考海报按以下维度评分，锚定评分标准。切勿自行选择、搜索或生成锚点；若无参考集，则不校准并标记`CALIBRATION: none`——切勿伪造锚点评分。维度（权重总和1.0）：设计0.35 · 工艺0.30 · 功能性0.20 · 原创性0.15。评分映射：`SCORE = min(round(1 + 9 × COMPOSITE), 触发的最低上限)`——上限在映射后应用，循环中的`Score ≥ 9`阈值始终读取最终的上限后SCORE，而非原始综合得分。

严格评分1–10。**关键上限**（严格下限——校准后的综合得分绝不能覆盖）：<2张真实论文图表 → ≤3；画布破损/内容裁剪/公式不可读 → ≤4；≥4个可见色相家族或渐变密集页眉 → ≤4；大面积空白卡片或列 → ≤5；伪造视觉论点 → ≤3。检查项：posterly展示级整体效果（该海报能否与专业设计的海报一同展示，而不会像拼凑的仪表盘？）、单强调色规范、真实图表清晰且居中、打印层级（标题 → 核心数据 → 图表 → 细节）、列填充、**公式突出度**（无超大框中的微小公式）、衬线正文/无衬线标题配对、无渐变俗套设计、组件一致性、60秒叙事性。输出格式：


循环：修复（使用以下修复词汇表）→ 重新执行第4阶段校验 → 重新评分。**每轮≤3个问题，最多3轮**。评分≥9 → 第6阶段。3轮后仍<9 → 停止补丁；升级为重新选择模板/画布/内容（回到第3阶段）或人工决策。绝不能进入第4轮 cosmetic补丁。

Fix vocabulary (closed set — the anti-patch-loop core)

修复词汇表（封闭集合——防补丁循环核心）

Allowed: (a) edit a

:root

token value; (b) swap/remove/add a whole component instance from

templates/COMPONENTS.md

; (c) content rebalance (move a card across columns, trim/grow text from the paper, resize a figure within its AR band); (d) template/canvas re-choice; (e) global edits to an existing component's CSS that reference only tokens; (f) switching predefined variants (

.eqn--large

.card--compact

.figure--wide

.nowrap

, …); (g) asset fixes (re-crop, swap for a clearer figure from the same paper, re-preprocess).

Forbidden: new inline styles, new hex values anywhere, bespoke decorative SVG, per-element font-size overrides. A new component may not be born inside the visual loop — stop, get a human checkpoint, add it to

COMPONENTS.md

, re-run from Phase 3.

允许：(a) 修改

:root

令牌值；(b) 从

templates/COMPONENTS.md

中替换/移除/添加完整组件实例；(c) 内容平衡（跨列移动卡片、从论文中删减/扩充文本、在宽高比范围内调整图表尺寸）；(d) 重新选择模板/画布；(e) 对现有组件的CSS进行全局编辑，且仅引用令牌；(f) 切换预定义变体（

.eqn--large

.card--compact

.figure--wide

.nowrap

, …）；(g) 资源修复（重新裁剪、替换为论文中更清晰的图表、重新预处理）。

禁止：新增内联样式、任何位置新增十六进制颜色值、定制装饰性SVG、逐元素字体大小覆盖。视觉循环中不得新增组件——停止流程，获取人工检查点，将组件添加到

COMPONENTS.md

，从第3阶段重新执行。

Phase 6 — Codex final review (fresh thread, cross-model)

第6阶段 —— Codex最终评审（全新线程，跨模型）

All hard gates PASS + polish warnings zero-or-waived + visual ≥ 9 first. Then a fresh codex thread (

xhigh

) reviews the final artifacts (not the content plan):

poster.html

, the rendered PDF/PNG, the paper source,

GATE_REPORT.json

CLAIM_EVIDENCE.md

— paths only, no executor framing. It checks: (1) fidelity & overclaims re-checked on final text (polish introduces new claims), (2) residue (

\ref{

TODO

, raw

in math, missing images, remote URLs), (3) visual rhetoric (headline numbers prominent, banner readable from 2 m), (4) gate-log coherence. The reviewer recommends; it does not edit. Any fix → back through Phase 4/5 gates — never straight to re-review.

所有严格校验通过 + 优化警告为零或已豁免 + 视觉评分≥9后，进入此阶段。使用全新codex线程（

xhigh

推理等级）评审最终产物（而非内容规划）：

poster.html

、渲染后的PDF/PNG、论文源文件、

GATE_REPORT.json

、

CLAIM_EVIDENCE.md

——仅传路径，无需执行框架说明。检查内容：(1) 最终文本的准确性与过度陈述重检（优化可能引入新论点），(2) 残留内容（

\ref{

TODO

, 数学公式中的原始

、缺失图像、远程URL），(3) 视觉修辞（核心数据突出、标题在2米外可读），(4) 校验逻辑一致性。评审者仅提供建议；不进行编辑。任何修复需回到第4/5阶段校验——绝不能直接进入重新评审。

Phase 7 — Final verification + report

第7阶段 —— 最终验证 + 报告

bash

python3 "$SKILL_SCRIPTS/poster_check.py" verify-final poster_html/poster_preview.pdf \
    --from-html poster_html/poster.html --max-size-mb 20

Page count 1, dimensions match

@page

, size ≤ 20 MB, no TODO/residue, no remote assets. Report: PDF path, final spread px, footer-gap range, gate summary table, unresolved waivers, codex verdict. Update

POSTER_STATE.json

→

done

bash

python3 "$SKILL_SCRIPTS/poster_check.py" verify-final poster_html/poster_preview.pdf \
    --from-html poster_html/poster.html --max-size-mb 20

页数为1，尺寸匹配

@page

设置，大小≤20 MB，无TODO/残留内容，无远程资源。报告内容：PDF路径、最终间距px、页脚间隙范围、校验摘要表、未解决的豁免项、codex评审结论。更新

POSTER_STATE.json

→

done

。

State persistence

状态持久化

poster_html/POSTER_STATE.json

{phase, venue, canvas{w,h,orientation,source_url, retrieved}, template, token_pack, design_decisions{...}, figures_selected[], visual_rounds, codex_threads{audit, final}, status, timestamp}

— written after every phase; enables compact-recovery resume.

poster_html/POSTER_STATE.json

：

{phase, venue, canvas{w,h,orientation,source_url,retrieved}, template, token_pack, design_decisions{...}, figures_selected[], visual_rounds, codex_threads{audit, final}, status, timestamp}

——每个阶段后写入；支持快速恢复流程。

Key rules

核心规则

Measure, don't eyeball. No layout claim without
```
run_gates.py
```
output.
Gates before aesthetics. Claude/codex review only ever sees a poster whose hard gates PASS. This ordering is what kills the patch-loop death spiral.
Never invent paper numbers or figures. Numbers come from the paper source; visuals carry manifest provenance. Fabrication = critical cap ≤ 3.
Two hues, one system. Accent + gold + neutrals. The style gate enforces it; don't negotiate with the gate.
Real figures are the poster. A poster without the paper's own figures is a dashboard, not a poster.
Fix vocabulary is closed. If a fix isn't expressible as token / component / rebalance / asset / canvas, it's the wrong fix.
Cross-model verdicts. Claude drives the loop and scores visuals; acceptance of content fidelity comes from the fresh codex thread (a loop can drive, never acquit).
Preserve user decisions. Re-read
```
design_decisions
```
before "improving" anything.
Vendor boundary.
```
poster_check.py
```
,
```
render_preview.py
```
,
```
_posterly/
```
are vendored from posterly — keep diffs minimal; ARIS-side logic goes in the new scripts, not in vendored files.

测量，而非目测。无
```
run_gates.py
```
输出则不得做出布局结论。
校验优先于审美。Claude/codex评审仅能看到严格校验通过的海报。此顺序可避免补丁循环的恶性循环。
切勿发明论文数据或图表。数据来自论文源文件；视觉内容需带来源清单。伪造内容→关键上限≤3。
双色调，单系统。强调色+金色+中性色。样式校验会强制执行；切勿与校验机制妥协。
真实图表是海报的核心。无论文自身图表的海报只是仪表盘，而非海报。
修复词汇表是封闭的。若修复无法用令牌/组件/平衡/资源/画布表达，则该修复是错误的。
跨模型结论。Claude驱动循环并进行视觉评分；内容准确性的认可来自全新codex线程（循环仅能驱动，绝不能自行确认）。
保留用户决策。进行任何“优化”前需重新读取
```
design_decisions
```
。
供应商边界。
```
poster_check.py
```
,
```
render_preview.py
```
,
```
_posterly/
```
来自posterly——尽量减少差异；ARIS侧逻辑放入新脚本，而非修改供应商文件。

Review tracing

评审追踪

Save every codex reviewer call's trace per

shared-references/review-tracing.md

.aris/traces/paper-poster-html/<date>_run<NN>/

(audit + final threads, raw responses).

按照

shared-references/review-tracing.md

保存每次codex评审调用的追踪记录至

.aris/traces/paper-poster-html/<date>_run<NN>/

（审核+最终线程，原始响应）。

Output contract

输出约定

poster_html/
├── poster.html              # single-file source of truth
├── poster_preview.pdf       # print-emulated, verify-final-checked
├── poster_preview.png       # thumbnail
├── POSTER_STATE.json        # resume state
├── GATE_REPORT.json         # canonical gate ledger (schema v1)
├── POSTER_CONTENT_PLAN.md   # what-goes-where + word budgets
├── CLAIM_EVIDENCE.md        # codex claim→evidence audit
├── FIGURE_MANIFEST.json     # figure provenance (sha256, page, bbox, dpi)
└── assets/{paper_figures,logos,qr,mathjax}/

poster_html/
├── poster.html              # 单文件可信源
├── poster_preview.pdf       # 打印模拟生成，已通过最终验证
├── poster_preview.png       # 缩略图
├── POSTER_STATE.json        # 恢复状态
├── GATE_REPORT.json         # 标准校验记录（schema v1）
├── POSTER_CONTENT_PLAN.md   # 内容分布+字数预算
├── CLAIM_EVIDENCE.md        # codex论点→证据审核
├── FIGURE_MANIFEST.json     # 图表来源信息（sha256，页码，边界框，DPI）
└── assets/{paper_figures,logos,qr,mathjax}/

When NOT to use

不适用场景

Slides, not a poster →
```
/paper-talk
```
/
```
/slides-polish
```
.
The paper's headline isn't stable yet — fix the paper first; a poster amplifies whatever story it's given.

需要幻灯片而非海报 → 使用
```
/paper-talk
```
/
```
/slides-polish
```
。
论文核心观点尚未稳定——先完善论文；海报只会放大它所承载的故事。