paper-talk

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Paper Talk: End-to-End Conference Talk Pipeline

Paper Talk:端到端会议演讲全流程

Workflow: from a completed paper to a conference-ready talk artifact — slide outline, Beamer source, editable PPTX, speaker notes, full talk script, polished visuals, and assurance checks.
Pipeline target: $ARGUMENTS
工作流:从已完成的论文到可直接用于会议的演讲成果——幻灯片大纲、Beamer源码、可编辑PPTX、演讲者备注、完整讲稿、优化后的视觉效果,以及合规检查。
流程目标:$ARGUMENTS

Assurance Ladder

合规等级阶梯

The skill takes an explicit
— assurance:
argument; default is
polished
.
LevelPhases runUse when
draft
0 → 1 → 2 → 5 → 6Internal practice talk; quick first-pass deck
polished
(default)
0 → 1 → 2 → 3 → 5 → 6Lab seminar, workshop, default conference talk
conference-ready
(or
submission
)
0 → 1 → 2 → 3 → 4 → 5 → 6Top-venue oral / spotlight where slide claims are scrutinised; or anonymous submission
Phase 4 (assurance checks) is opt-in via
— assurance: conference-ready
. The lower levels skip claim / citation / anonymity audits, which is correct when content has already been audited at the paper-writing stage.
该技能支持显式的
— assurance:
参数;默认值为
polished
等级执行阶段适用场景
draft
0 → 1 → 2 → 5 → 6内部演练演讲;快速生成第一版幻灯片
polished
(默认)
0 → 1 → 2 → 3 → 5 → 6实验室研讨会、工作坊、常规会议演讲
conference-ready
(或
submission
0 → 1 → 2 → 3 → 4 → 5 → 6顶级会议口头报告/重点展示(幻灯片论点会被严格审查);或匿名投稿演讲
阶段4(合规检查)需通过
— assurance: conference-ready
主动开启。 低等级会跳过论点/引用/匿名性审计,这在论文撰写阶段已完成内容审核的场景下是合理的。

Hard Invariants

核心不变规则

These are non-negotiable across all phases:
  1. Original paper is read-only. This workflow consumes the paper directory; it never modifies
    paper/main.tex
    or other paper artifacts.
  2. Original deck is preserved.
    /paper-slides
    produces a baseline deck;
    /slides-polish
    writes a
    _polished
    versioned copy. The original Beamer + PPTX are never overwritten.
  3. Speaker notes are byte-stable. Polish must not change
    slide.notes_slide
    content. Phase 4 verifies this.
  4. No new content anywhere in the pipeline. All slide text, speaker notes, talk script, Q&A answers, claims, numbers, citations, URLs, author names, affiliations, anonymity placeholders, and experiment results must be either paper-grounded (extracted from
    PAPER_DIR/
    artefacts) or explicitly user-provided. The pipeline never invents content during outline, build, polish, audit, or export. Phase-4 anonymity scan + claim audit verify this end-to-end.
  5. No slide reordering. Add / drop / reorder requires explicit user flags.
  6. Cross-model independence. Per-page Codex calls in
    /slides-polish
    use fresh threads (no
    codex-reply
    ). See
    ../shared-references/reviewer-independence.md
    .
  7. Anonymity fail-closed. If any audit (or any Codex fix proposal) would replace a placeholder with a real title / count / URL, the workflow halts and surfaces the proposal for human review. See
    ../shared-references/experiment-integrity.md
    .
  8. Style references are guidance, not text source. A
    — reference:
    PDF or
    — style:
    preset informs visual weight and structural rhythm; never copy prose, examples, slide titles, or speaker-note text from the reference.
  9. Final report cannot be
    conference-ready
    unless required audits pass.
    Phase 6 verifies and downgrades verdict if audits fail.
  10. reasoning_effort: xhigh
    is invariant across all
    effort
    levels for any Codex call invoked by sub-skills.
以下规则在所有阶段均不可违反:
  1. 原始论文为只读状态。该工作流仅读取论文目录;绝不会修改
    paper/main.tex
    或其他论文产物。
  2. 原始幻灯片保留
    /paper-slides
    生成基线幻灯片;
    /slides-polish
    会写入带
    _polished
    后缀的版本副本。原始Beamer + PPTX文件绝不会被覆盖。
  3. 演讲者备注字节级稳定。优化阶段不得修改
    slide.notes_slide
    内容。阶段4会对此进行验证。
  4. 流程中不得生成新内容。所有幻灯片文本、演讲者备注、讲稿、问答内容、论点、数据、引用、URL、作者姓名、机构信息、匿名占位符及实验结果,必须要么来自论文(从
    PAPER_DIR/
    产物中提取),要么由用户明确提供。在大纲生成、幻灯片构建、优化、审计或导出阶段,流程绝不会自行生成内容。阶段4的匿名性扫描+论点审计会端到端验证这一点。
  5. 不得重新排序幻灯片。添加/删除/重新排序需用户明确设置参数。
  6. 跨模型独立
    /slides-polish
    中的逐页Codex调用使用全新线程(不依赖
    codex-reply
    )。详见
    ../shared-references/reviewer-independence.md
  7. 匿名性检查失败即终止。若任何审计(或Codex修复建议)会将占位符替换为真实标题/数据/URL,工作流将暂停并将建议提交给人工审核。详见
    ../shared-references/experiment-integrity.md
  8. 风格参考仅作指导,而非文本来源
    — reference:
    指定的PDF或
    — style:
    预设仅用于指导视觉权重和结构节奏;绝不能从参考中复制文案、示例、幻灯片标题或演讲者备注文本。
  9. 最终报告仅在通过要求的审计后才能标记为
    conference-ready
    。阶段6会进行验证,若审计失败则降级结论。
  10. **
    reasoning_effort: xhigh
    **在所有
    effort
    等级下,对所有子技能调用的Codex请求均保持不变。

Constants

固定参数

  • PAPER_DIR =
    paper/
    — Source paper directory. Override via positional argument.
  • OUTPUT_DIR =
    slides/
    — Where the deck artefacts live.
  • STATE_DIR =
    .aris/paper-talk/
    — Workflow state, audit logs, final report.
  • TALK_TYPE =
    spotlight
    — Default talk format. Inherited by
    /paper-slides
    .
  • TALK_MINUTES = 15 — Default duration. Inherited by
    /paper-slides
    .
  • VENUE =
    NeurIPS
    — Default venue (used by
    /paper-slides
    color schemes when
    — style
    is not passed).
  • ASPECT_RATIO =
    16:9
    — Inherited by
    /paper-slides
    .
  • STYLE_PRESET
    generic
    if not passed;
    why-rf
    and venue presets supported by
    /slides-polish
    .
  • REFERENCE_VISUAL — Required when
    assurance ≥ polished
    . The Beamer compile of this talk is an acceptable self-reference; an external academic talk PDF is preferred when the user wants visual alignment beyond defaults.
  • AUTO_PROCEED = false — Each major phase pauses for user approval. Set
    true
    only when explicitly requested.
  • PAPER_DIR =
    paper/
    — 源论文目录。可通过位置参数覆盖。
  • OUTPUT_DIR =
    slides/
    — 幻灯片产物存储目录。
  • STATE_DIR =
    .aris/paper-talk/
    — 工作流状态、审计日志、最终报告存储目录。
  • TALK_TYPE =
    spotlight
    — 默认演讲格式。由
    /paper-slides
    继承。
  • TALK_MINUTES = 15 — 默认演讲时长。由
    /paper-slides
    继承。
  • VENUE =
    NeurIPS
    — 默认会议场地(当未传递
    — style
    参数时,
    /paper-slides
    会使用该场地的配色方案)。
  • ASPECT_RATIO =
    16:9
    — 由
    /paper-slides
    继承。
  • STYLE_PRESET — 未传递时为
    generic
    /slides-polish
    支持
    why-rf
    及场地预设。
  • REFERENCE_VISUAL — 当
    assurance ≥ polished
    时为必填项。该演讲的Beamer编译版可作为自参考;若用户希望视觉效果超出默认值,优先使用外部学术演讲PDF。
  • AUTO_PROCEED = false — 每个主要阶段都会暂停等待用户确认。仅当用户明确要求时才设置为
    true

Inputs

输入项

Discovered from
$ARGUMENTS
and the project directory:
  1. PAPER_DIR/
    — compiled paper (
    main.tex
    +
    main.pdf
    ). The paper must already be in good shape; this workflow does not write the paper.
  2. PAPER_DIR/sections/*.tex
    — section sources for content extraction.
  3. PAPER_DIR/figures/
    — figures available for slide reuse.
  4. (optional)
    slides/SLIDE_OUTLINE.md
    — pre-existing outline (resume mode); skips Phase 1 generation.
  5. (optional)
    — reference: <pdf>
    — visual anchor for
    /slides-polish
    . If absent and the assurance level requires polish, the workflow uses the Beamer self-compile as reference.
  6. (optional)
    — talk_type
    ,
    — minutes
    ,
    — style
    ,
    — effort
    — see Constants.
$ARGUMENTS
和项目目录中获取:
  1. PAPER_DIR/
    — 已编译的论文(
    main.tex
    +
    main.pdf
    )。论文需已完成定稿;该工作流不负责撰写论文。
  2. PAPER_DIR/sections/*.tex
    — 用于内容提取的章节源码。
  3. PAPER_DIR/figures/
    — 可复用至幻灯片的图片。
  4. (可选)
    slides/SLIDE_OUTLINE.md
    — 已存在的大纲(恢复模式);跳过阶段1的生成步骤。
  5. (可选)
    — reference: <pdf>
    /slides-polish
    的视觉锚点。若未提供且合规等级要求优化,工作流会使用Beamer自编译版作为参考。
  6. (可选)
    — talk_type
    ,
    — minutes
    ,
    — style
    ,
    — effort
    — 详见固定参数。

Output Layout

输出结构

File names match
/paper-slides
's emit contract — this workflow is a non-renaming consumer.
slides/
├── SLIDE_OUTLINE.md                # Phase-1 outline (claim-first per slide)
├── main.tex                        # Beamer source (from /paper-slides)
├── main.pdf                        # compiled Beamer
├── presentation.pptx               # editable PPTX (from /paper-slides; emit name fixed by /paper-slides)
├── presentation_pre_polish.pptx    # snapshot before /slides-polish
├── presentation_polished.pptx      # /slides-polish output (when assurance ≥ polished)
├── presentation_polished.pdf       # rendered (LibreOffice or user-export)
├── speaker_notes.md                # per-slide notes (from /paper-slides)
├── TALK_SCRIPT.md                  # full word-for-word talk script + Q&A (from /paper-slides)
└── assets/                         # per-slide PNG previews

.aris/paper-talk/
├── PIPELINE_STATE.json             # phase pointer, status, timestamps
├── FINAL_REPORT.md                 # human-readable summary at end
├── audit-input/                    # Phase-4 staging copies of slide text + notes + script (so /paper-claim-audit and /citation-audit can run on slide content as a synthetic "paper")
│   ├── main.tex                    # \input{sections/slide_text.tex} \input{sections/notes.tex} \input{sections/script.tex}
│   ├── sections/
│   │   ├── slide_text.tex          # all visible slide bullets/titles/callouts as \section{Slide K}
│   │   ├── notes.tex               # speaker_notes.md → LaTeX, one \section per slide
│   │   └── script.tex              # TALK_SCRIPT.md → LaTeX, one \section per slide
│   ├── references.bib → ../../../paper/references.bib   # symlink to source paper bib for /citation-audit
│   ├── results/ → ../../../paper/results/                # symlink — claim audit needs real evidence
│   └── figures/ → ../../../paper/figures/                # symlink — citation audit may reference figure captions
└── audits/
    ├── slide_claim_audit.json      # 6-state verdict per claim (PASS/WARN/FAIL/NOT_APPLICABLE/BLOCKED/ERROR)
    ├── citation_audit.json         # citation existence / metadata
    ├── anonymity_scan.json         # placeholder discipline check
    └── export_integrity.json       # page-count / aspect / blank-slide / notes-preserved
The audit JSON files follow the shared 6-state schema; see
../shared-references/assurance-contract.md
.
文件名遵循
/paper-slides
的输出约定 — 该工作流为非重命名消费者
slides/
├── SLIDE_OUTLINE.md                # 阶段1生成的大纲(每页以论点为核心)
├── main.tex                        # Beamer源码(来自/paper-slides)
├── main.pdf                        # 编译后的Beamer文件
├── presentation.pptx               # 可编辑PPTX(来自/paper-slides;文件名由/paper-slides固定)
├── presentation_pre_polish.pptx    # /slides-polish优化前的快照
├── presentation_polished.pptx      # /slides-polish输出文件(当assurance ≥ polished时生成)
├── presentation_polished.pdf       # 渲染后的文件(LibreOffice或用户导出)
├── speaker_notes.md                # 逐页演讲者备注(来自/paper-slides)
├── TALK_SCRIPT.md                  # 完整逐字讲稿+问答内容(来自/paper-slides)
└── assets/                         # 逐页PNG预览图

.aris/paper-talk/
├── PIPELINE_STATE.json             # 阶段指针、状态、时间戳
├── FINAL_REPORT.md                 # 最终可读总结报告
├── audit-input/                    # 阶段4的幻灯片文本+备注+讲稿暂存副本(以便/paper-claim-audit和/citation-audit将幻灯片内容作为合成"论文"进行审核)
│   ├── main.tex                    # \\input{sections/slide_text.tex} \\input{sections/notes.tex} \\input{sections/script.tex}
│   ├── sections/
│   │   ├── slide_text.tex          # 所有可见幻灯片项目符号/标题/标注,按\\section{Slide K}组织
│   │   ├── notes.tex               # speaker_notes.md转换为LaTeX格式,每页对应一个\\section
│   │   └── script.tex              # TALK_SCRIPT.md转换为LaTeX格式,每页对应一个\\section
│   ├── references.bib → ../../../paper/references.bib   # 指向源论文bib文件的符号链接
│   ├── results/ → ../../../paper/results/                # 符号链接 — 论点审核需要真实证据
│   └── figures/ → ../../../paper/figures/                # 符号链接 — 引用审核可能需要参考图片标题
└── audits/
    ├── slide_claim_audit.json      # 每个论点的6种状态结论(PASS/WARN/FAIL/NOT_APPLICABLE/BLOCKED/ERROR)
    ├── citation_audit.json         # 引用存在性/元数据检查
    ├── anonymity_scan.json         # 占位符规范性检查
    └── export_integrity.json       # 页数/比例/空白幻灯片/备注保留情况检查
审计JSON文件遵循通用的6状态 schema;详见
../shared-references/assurance-contract.md

Workflow

工作流

Phase 0: Setup

阶段0:初始化

  1. Validate paper directory:
    ls $PAPER_DIR/main.tex $PAPER_DIR/main.pdf
    . If absent → halt and ask user to point to the compiled paper.
  2. Check prerequisites:
    • LaTeX:
      which xelatex pdflatex latexmk
    • PPTX rendering:
      which soffice
      (LibreOffice headless) — required for Phase 4 export integrity check; otherwise prompt user to export PDF manually.
    • PDF tools:
      which pdfinfo pdftoppm
      (poppler) — required for
      /slides-polish
      PNG rendering.
    • Codex MCP availability.
    • python-pptx (
      python3 -c 'import pptx'
      ).
  3. Resolve overrides from
    $ARGUMENTS
    :
    talk_type
    ,
    minutes
    ,
    assurance
    ,
    reference
    ,
    style
    ,
    effort
    .
  4. State init: write
    .aris/paper-talk/PIPELINE_STATE.json
    with
    phase: 0
    , timestamp, all resolved overrides.
  5. Resume mode: if
    slides/SLIDE_OUTLINE.md
    exists and
    PIPELINE_STATE.json
    shows recent in-progress work, prompt the user — resume from last phase or start fresh.
  1. 验证论文目录:检查
    $PAPER_DIR/main.tex $PAPER_DIR/main.pdf
    是否存在。若缺失→暂停并提示用户指定已编译的论文。
  2. 检查前置依赖
    • LaTeX:
      which xelatex pdflatex latexmk
    • PPTX渲染:
      which soffice
      (LibreOffice无头模式)—— 阶段4导出完整性检查必需;否则提示用户手动导出PDF。
    • PDF工具:
      which pdfinfo pdftoppm
      (poppler)——
      /slides-polish
      生成PNG预览必需。
    • Codex MCP可用性。
    • python-pptx(执行
      python3 -c 'import pptx'
      验证)。
  3. 解析
    $ARGUMENTS
    中的覆盖参数
    talk_type
    ,
    minutes
    ,
    assurance
    ,
    reference
    ,
    style
    ,
    effort
  4. 初始化状态:写入
    .aris/paper-talk/PIPELINE_STATE.json
    ,包含
    phase: 0
    、时间戳及所有解析后的覆盖参数。
  5. 恢复模式:若
    slides/SLIDE_OUTLINE.md
    存在且
    PIPELINE_STATE.json
    显示近期有未完成工作,提示用户——从上次阶段恢复或重新开始。

Phase 1: Slide Outline (Checkpoint)

阶段1:幻灯片大纲(检查点)

Goal: produce or accept a
slides/SLIDE_OUTLINE.md
that the user signs off on before any deck-building happens.
If
slides/SLIDE_OUTLINE.md
already exists, present its summary (slide count, time budget, claim-per-slide map) to the user and ask:
"Use existing outline (Y/n)? Modify? Regenerate?"
Otherwise, delegate to
/paper-slides
Phase-1 only (content extraction + slide outline generation).
/paper-slides
writes the outline into its own state; we adopt it as
slides/SLIDE_OUTLINE.md
.
The outline must contain, per slide:
  • A claim-first title (states the slide's point, not the topic).
  • One main idea per slide.
  • Bullet content (sentence fragments, not paragraphs).
  • Figure / equation / callout note.
  • Time budget (seconds).
  • Transition cue.
  • Speaker note seed (1-3 sentences; expanded in Phase 2).
Checkpoint: present the outline. Default behavior: pause for user approval. Set
AUTO_PROCEED = true
only when the user is explicitly running unattended.
目标:生成或接收
slides/SLIDE_OUTLINE.md
,并在开始构建幻灯片前获得用户确认。
slides/SLIDE_OUTLINE.md
已存在,向用户展示其摘要(幻灯片数量、时间分配、每页论点映射)并询问:
"是否使用现有大纲(Y/n)?修改?重新生成?"
否则,仅调用
/paper-slides
的阶段1(内容提取+幻灯片大纲生成)。
/paper-slides
会将大纲写入自身状态;我们将其复制为
slides/SLIDE_OUTLINE.md
大纲需包含每页幻灯片的以下内容:
  • 以论点为核心的标题(明确幻灯片的核心观点,而非仅主题)。
  • 每页一个核心思想。
  • 项目符号内容(短句片段,而非段落)。
  • 图片/公式/标注说明。
  • 时间分配(秒)。
  • 过渡提示。
  • 演讲者备注初稿(1-3句话;阶段2会扩展)。
检查点:展示大纲。默认行为:暂停等待用户确认。仅当用户明确要求无人值守时设置
AUTO_PROCEED = true

Phase 2: Build Baseline Deck

阶段2:构建基线幻灯片

Invoke
/paper-slides
to generate Beamer source + PPTX from the approved outline.
/paper-slides "<paper-dir>" — talk_type: <T> — minutes: <N> — venue: <V> — aspect: 16:9 — notes: true
Forward
— style-ref:
if the user originally passed it (writer-side; the reviewer-side polish in Phase 3 must NOT see it — see invariant 8).
/paper-slides
produces (emit names fixed by
/paper-slides
):
  • slides/main.tex
    +
    slides/main.pdf
    (Beamer)
  • slides/presentation.pptx
    (editable PPTX)
  • slides/speaker_notes.md
  • slides/TALK_SCRIPT.md
After return: verify the four artefacts exist. If any missing, halt with a diagnostic. Do not rename outputs — downstream phases bind to these exact paths.
调用
/paper-slides
根据已确认的大纲生成Beamer源码+PPTX。
/paper-slides "<paper-dir>" — talk_type: <T> — minutes: <N> — venue: <V> — aspect: 16:9 — notes: true
若用户最初传递了
— style-ref:
参数,则转发该参数(仅用于撰写侧;阶段3的审核侧优化不得使用该参数——见核心不变规则8)。
/paper-slides
生成以下产物(文件名由
/paper-slides
固定):
  • slides/main.tex
    +
    slides/main.pdf
    (Beamer文件)
  • slides/presentation.pptx
    (可编辑PPTX)
  • slides/speaker_notes.md
  • slides/TALK_SCRIPT.md
返回后:验证这四个产物是否存在。若有缺失,暂停并输出诊断信息。不得重命名输出——后续阶段依赖这些精确路径。

Phase 3: Visual Polish (when assurance ≥ polished)

阶段3:视觉优化(当assurance ≥ polished时执行)

Skip when
assurance == draft
.
Invoke
/slides-polish
against the freshly generated PPTX, with a reference. If the user passed
— reference:
, use that. Otherwise, use the Beamer compile (
slides/main.pdf
) as the visual reference — the Beamer is the design intent for the same talk, so it is the correct anchor when no external reference is available.
/slides-polish "slides/presentation.pptx" — reference: slides/main.pdf — style: <preset> — effort: <effort>
After return: verify
slides/presentation_polished.pptx
exists; if rendering tools are available, also verify
slides/presentation_polished.pdf
.
The polish phase is read-only on content (see Hard Invariants). If
/slides-polish
emits a
BLOCKED
verdict because a Codex fix proposal would alter content, surface that block to the user and halt rather than overriding.
assurance == draft
时跳过。
调用
/slides-polish
对刚生成的PPTX进行优化,并指定参考文件。若用户传递了
— reference:
,则使用该文件;否则使用Beamer编译版(
slides/main.pdf
)作为视觉参考——Beamer是同一演讲的设计意图,因此在无外部参考时是正确的锚点。
/slides-polish "slides/presentation.pptx" — reference: slides/main.pdf — style: <preset> — effort: <effort>
返回后:验证
slides/presentation_polished.pptx
是否存在;若渲染工具可用,同时验证
slides/presentation_polished.pdf
是否存在。
优化阶段对内容为只读(见核心不变规则)。若
/slides-polish
因Codex修复建议会修改内容而输出
BLOCKED
结论,需将该阻塞提示给用户并暂停,而非强制覆盖。

Phase 4: Assurance Checks (when assurance == conference-ready)

阶段4:合规检查(当assurance == conference-ready时执行)

Skip when
assurance ∈ {draft, polished}
.
/paper-claim-audit
and
/citation-audit
expect a paper-shaped input (
paper/main.tex
+
paper/sections/*.tex
+
paper/results/
) and emit JSON under
paper/
. Slides have a different shape, so this phase first stages slide artefacts into a synthetic paper directory, then invokes the audits against that synthetic input. Verdicts are written under
.aris/paper-talk/audits/
using the shared 6-state schema (
PASS
,
WARN
,
FAIL
,
NOT_APPLICABLE
,
BLOCKED
,
ERROR
) — see
../shared-references/assurance-contract.md
.
assurance ∈ {draft, polished}
时跳过。
/paper-claim-audit
/citation-audit
期望论文格式的输入(
paper/main.tex
+
paper/sections/*.tex
+
paper/results/
)并在
paper/
目录下输出JSON。幻灯片格式不同,因此该阶段首先将幻灯片产物暂存为合成论文目录,然后针对该合成输入执行审核。结论将写入
.aris/paper-talk/audits/
,使用通用的6状态 schema(
PASS
,
WARN
,
FAIL
,
NOT_APPLICABLE
,
BLOCKED
,
ERROR
)——详见
../shared-references/assurance-contract.md

4.0 Staging adapter

4.0 暂存适配

The staged "paper" mirrors
/paper-claim-audit
and
/citation-audit
's expected layout:
main.tex
at the root that
\input{}
s
sections/*.tex
, plus a
.bib
file and
results/
+
figures/
symlinks.
.aris/paper-talk/audit-input/
├── main.tex
│   # contains: \input{sections/slide_text.tex}
│   #           \input{sections/notes.tex}
│   #           \input{sections/script.tex}
│   #           \bibliography{references}
├── sections/
│   ├── slide_text.tex      # one \section{Slide K — <title>} per slide; visible bullets, titles, callout text, in-slide \cite{...}
│   ├── notes.tex           # speaker_notes.md converted to LaTeX, one \section per slide
│   └── script.tex          # TALK_SCRIPT.md converted to LaTeX, one \section per slide
├── references.bib → ../../../paper/references.bib  # symlink to source paper bib (or copy if symlink not allowed)
├── results/ → ../../../paper/results/              # symlink — claim audit needs real evidence
└── figures/ → ../../../paper/figures/              # symlink — citation audit may reference figure captions
If the source paper uses a different bib name (e.g.,
paper.bib
), mirror the actual file. If multiple bibs are used, link them all.
The staged "paper" exists only for the duration of Phase 4 and is removed in Phase 6 unless the user passes
— keep-audit-input
.
暂存的"论文"镜像
/paper-claim-audit
/citation-audit
期望的目录结构:根目录下的
main.tex
通过
\\input{}
引入
sections/*.tex
,加上
.bib
文件及
results/
+
figures/
符号链接。
.aris/paper-talk/audit-input/
├── main.tex
│   # 内容:\\input{sections/slide_text.tex}
│   #           \\input{sections/notes.tex}
│   #           \\input{sections/script.tex}
│   #           \\bibliography{references}
├── sections/
│   ├── slide_text.tex      # 每页对应一个\\section{Slide K — <title>};包含可见项目符号、标题、标注文本及幻灯片内的\\cite{...}
│   ├── notes.tex           # speaker_notes.md转换为LaTeX格式,每页对应一个\\section
│   └── script.tex          # TALK_SCRIPT.md转换为LaTeX格式,每页对应一个\\section
├── references.bib → ../../../paper/references.bib  # 指向源论文bib文件的符号链接(若不允许符号链接则复制)
├── results/ → ../../../paper/results/              # 符号链接 — 论点审核需要真实证据
└── figures/ → ../../../paper/figures/              # 符号链接 — 引用审核可能需要参考图片标题
若源论文使用不同的bib文件名(如
paper.bib
),则镜像实际文件。若使用多个bib文件,则全部链接。
暂存的"论文"仅在阶段4存在,阶段6会删除,除非用户传递
— keep-audit-input
参数。

4.1 Slide claim audit

4.1 幻灯片论点审核

Invoke
/paper-claim-audit
against the staged input. Scope is slide text + speaker notes + full talk script — talks often smuggle unsupported claims into spoken parts that the visible bullets don't show.
/paper-claim-audit ".aris/paper-talk/audit-input"
The audit emits
audit-input/PAPER_CLAIM_AUDIT.json
with the shared 6-state schema. Move it to
.aris/paper-talk/audits/slide_claim_audit.json
and de-stage the path prefix so the verdicts cite slide K rather than synthetic-paper sections.
A
FAIL
or
BLOCKED
verdict on any claim downgrades the Phase-6 final verdict from
conference-ready
to
polished
.
调用
/paper-claim-audit
针对暂存输入执行审核。范围覆盖幻灯片文本+演讲者备注+完整讲稿——演讲中常将未验证的论点隐藏在口头内容中,而幻灯片可见部分并未体现。
/paper-claim-audit ".aris/paper-talk/audit-input"
审核会在
audit-input/PAPER_CLAIM_AUDIT.json
中输出通用6状态结论。将其移动至
.aris/paper-talk/audits/slide_claim_audit.json
,并移除路径前缀,使结论指向第K页幻灯片而非合成论文章节。
若任何论点的结论为
FAIL
BLOCKED
,阶段6的最终结论将从
conference-ready
降级为
polished

4.2 Citation audit

4.2 引用审核

Invoke
/citation-audit
over the staged input. Verify any
\cite{...}
in slides + notes + script via DBLP / CrossRef; flag fabricated entries.
/citation-audit ".aris/paper-talk/audit-input"
Output →
.aris/paper-talk/audits/citation_audit.json
(6-state).
调用
/citation-audit
针对暂存输入执行审核。通过DBLP / CrossRef验证幻灯片+备注+讲稿中的所有
\\cite{...}
;标记伪造的引用条目。
/citation-audit ".aris/paper-talk/audit-input"
输出→
.aris/paper-talk/audits/citation_audit.json
(6状态)。

4.3 Anonymity scan

4.3 匿名性扫描

When the talk is for an anonymous-submission venue or the user passed
— anonymous
, scan slides + notes + script for:
  • Real paper titles where placeholders should be (
    [anonymized]
    ,
    Withheld for anonymous review
    , etc.).
  • Real author names beyond the talk's own author block.
  • Real submission counts.
  • Real URLs that would deanonymize.
Output →
.aris/paper-talk/audits/anonymity_scan.json
(6-state). Any
FAIL
(real-content leak) blocks
conference-ready
.
当演讲用于匿名投稿场地或用户传递
— anonymous
参数时,扫描幻灯片+备注+讲稿以检查:
  • 应使用占位符(
    [anonymized]
    Withheld for anonymous review
    等)的位置出现真实论文标题。
  • 演讲作者信息之外出现真实作者姓名。
  • 真实投稿数量。
  • 会暴露身份的真实URL。
输出→
.aris/paper-talk/audits/anonymity_scan.json
(6状态)。任何
FAIL
(真实内容泄露)会阻止结论标记为
conference-ready

Phase 5: Final Export + Integrity

阶段5:最终导出+完整性检查

Resolve the final PPTX path:
FINAL_PPTX = slides/presentation_polished.pptx  if Phase 3 ran
           = slides/presentation.pptx           if assurance == draft (Phase 3 skipped)
Then:
  1. Recompile Beamer cleanly:
    cd slides && latexmk -pdf -xelatex main.tex
    (or
    pdflatex
    for non-CJK). Confirm
    main.pdf
    page count matches outline slide count.
  2. Render
    FINAL_PPTX
    → PDF
    via
    soffice --headless --convert-to pdf <FINAL_PPTX> -outdir slides/
    . If unavailable, prompt user.
  3. Export integrity check
    .aris/paper-talk/audits/export_integrity.json
    (6-state):
    • PPTX-PDF page count == Beamer-PDF page count.
    • Aspect ratio == declared (16:9 default).
    • No fully-blank slide.
    • Speaker notes preserved on every slide that had them in the baseline. When Phase 3 ran, sha256 match against
      presentation_pre_polish.pptx
      's notes; when Phase 3 skipped, against
      presentation.pptx
      's own notes (no-op).
    • Embedded fonts are appropriate for projector use (e.g., Calibri / Helvetica Neue / PingFang SC for Chinese).
If any integrity check fails, downgrade the Phase-6 verdict and surface specifics to the user.
确定最终PPTX路径
FINAL_PPTX = slides/presentation_polished.pptx  若阶段3已执行
           = slides/presentation.pptx           若assurance == draft(阶段3跳过)
然后执行:
  1. 重新编译Beamer
    cd slides && latexmk -pdf -xelatex main.tex
    (非CJK文档使用
    pdflatex
    )。确认
    main.pdf
    页数与大纲幻灯片数量一致。
  2. 将FINAL_PPTX渲染为PDF:通过
    soffice --headless --convert-to pdf <FINAL_PPTX> -outdir slides/
    执行。若工具不可用,提示用户手动操作。
  3. 导出完整性检查
    .aris/paper-talk/audits/export_integrity.json
    (6状态):
    • PPTX导出PDF的页数 == Beamer PDF的页数。
    • 比例 == 声明的比例(默认16:9)。
    • 无完全空白的幻灯片。
    • 所有基线中包含备注的幻灯片,备注均已保留。若阶段3已执行,需与
      presentation_pre_polish.pptx
      的备注进行sha256匹配;若阶段3跳过,则与
      presentation.pptx
      自身的备注匹配(无操作)。
    • 嵌入字体适合投影使用(如中文使用Calibri / Helvetica Neue / PingFang SC)。
若任何完整性检查失败,阶段6的结论将降级,并向用户展示具体失败细节。

Phase 6: Final Report

阶段6:最终报告

Write
.aris/paper-talk/FINAL_REPORT.md
with:
  • Verdict:
    draft
    |
    polished
    |
    conference-ready
    (downgraded if audits failed).
  • Artefact paths: Beamer source / PDF, baseline PPTX, polished PPTX, polished PDF, notes, script.
  • Slide count, time budget vs target.
  • Audit summary (one line per audit run + PASS / WARN / FAIL).
  • Open warnings: any unresolved Codex polish notes the user should review by hand.
  • Next steps: e.g., "drop in real QR images on slide N", "verify HF Daily Papers screenshot", "rehearse to confirm 25-min budget".
A final
conference-ready
verdict requires:
  • Phases 0-5 all completed without halt.
  • All Phase-4 audit JSONs returned
    PASS
    (or
    NOT_APPLICABLE
    when the audit's content detector did not match — e.g., no citations on slides). Any
    WARN
    ,
    FAIL
    ,
    BLOCKED
    , or
    ERROR
    downgrades the verdict.
  • Phase-5 export integrity returned
    PASS
    .
  • No anonymity-scan
    FAIL
    when anonymous submission is declared.
Otherwise the report emits the highest passing level (
polished
or
draft
) and itemises why
conference-ready
was not granted, citing the specific audit JSON and the failing verdict line.
写入
.aris/paper-talk/FINAL_REPORT.md
,包含:
  • 结论
    draft
    |
    polished
    |
    conference-ready
    (若审计失败则降级)。
  • 产物路径:Beamer源码/PDF、基线PPTX、优化后PPTX、优化后PDF、备注、讲稿。
  • 幻灯片数量、时间分配与目标对比。
  • 审计总结(每次审核一行 + PASS / WARN / FAIL)。
  • 未解决警告:用户需手动查看的Codex优化未决建议。
  • 后续步骤:例如"在第N页插入真实二维码图片"、"验证HF Daily Papers截图"、"排练以确认25分钟时长"。
最终
conference-ready
结论需满足:
  • 阶段0-5全部完成且未暂停。
  • 阶段4所有审计JSON均返回
    PASS
    (或当审计内容检测器未匹配时返回
    NOT_APPLICABLE
    ——如幻灯片无引用)。任何
    WARN
    FAIL
    BLOCKED
    ERROR
    都会降级结论。
  • 阶段5导出完整性检查返回
    PASS
  • 当声明为匿名投稿时,匿名性扫描无
    FAIL
    结论。
否则报告将输出最高通过等级(
polished
draft
),并逐项说明未达到
conference-ready
的原因,引用具体审计JSON及失败结论行。

Effort Levels

工作量等级

See
../shared-references/effort-contract.md
.
effort
Behavior
lite
Forwarded to sub-skills.
/slides-polish
runs in
lite
(BLOCKERS only). Phase 4 still runs at the requested assurance level
effort
controls depth, never audit gating (see
../shared-references/assurance-contract.md
).
balanced
(default)
Standard pipeline.
/slides-polish
runs
balanced
.
max
/slides-polish
runs
max
(per-page review on every slide). Phase-4 audits read all artefacts in full.
beast
/slides-polish
runs
beast
(second polish round). Phase-4 audits include extended assurance checks; final report adds an executive summary.
reasoning_effort: xhigh
is invariant.
assurance
and
effort
are independent axes. The user may legally combine
effort: lite, assurance: conference-ready
to mean "fast pipeline, but every audit must emit a verdict before finalisation."
详见
../shared-references/effort-contract.md
effort
行为
lite
传递给子技能。
/slides-polish
lite
模式运行(仅处理阻塞项)。阶段4仍会按请求的合规等级执行——
effort
控制深度,绝不跳过审计(详见
../shared-references/assurance-contract.md
)。
balanced
(默认)
标准流程。
/slides-polish
balanced
模式运行。
max
/slides-polish
max
模式运行(逐页审核所有幻灯片)。阶段4审计会完整读取所有产物。
beast
/slides-polish
beast
模式运行(第二轮优化)。阶段4审计包含扩展合规检查;最终报告增加执行摘要。
reasoning_effort: xhigh
保持不变。
assurance
effort
独立维度。用户可合法组合
effort: lite, assurance: conference-ready
,意为"快速流程,但所有审计必须输出结论后才能完成最终定稿"。

Parameter Pass-Through

参数传递示例

/paper-talk "paper/" — talk_type: oral — minutes: 25 — assurance: conference-ready \
            — reference: ./refs/why_rf_2025.pdf — style: why-rf — effort: max
Forwarded to:
  • /paper-slides
    (Phase 2):
    paper-dir
    ,
    talk_type
    ,
    minutes
    ,
    venue
    ,
    aspect
    ,
    notes
    ,
    style-ref
    if writer-side.
  • /slides-polish
    (Phase 3): polished pptx target,
    reference
    ,
    style
    ,
    effort
    .
  • /paper-claim-audit
    (Phase 4.1): scoped to slide deck artefacts (not paper).
  • /citation-audit
    (Phase 4.2): scoped to slide-deck-only citations.
  • Anonymity scan (Phase 4.3): inline tool, no external skill.
/paper-talk "paper/" — talk_type: oral — minutes: 25 — assurance: conference-ready \\
            — reference: ./refs/why_rf_2025.pdf — style: why-rf — effort: max
传递给:
  • /paper-slides
    (阶段2):
    paper-dir
    ,
    talk_type
    ,
    minutes
    ,
    venue
    ,
    aspect
    ,
    notes
    , 若为撰写侧则传递
    style-ref
  • /slides-polish
    (阶段3):优化后的PPTX目标、
    reference
    ,
    style
    ,
    effort
  • /paper-claim-audit
    (阶段4.1):范围限定为幻灯片产物(而非论文)。
  • /citation-audit
    (阶段4.2):范围限定为幻灯片中的引用。
  • 匿名性扫描(阶段4.3):内置工具,无需调用外部技能。

When NOT to Use

不适用于以下场景

  • The paper is not yet compiled or claims are not stable. Run
    /paper-writing
    first.
  • The user wants a poster, not a talk. Use
    /paper-poster
    .
  • The user already has a finished deck and only needs visual polish. Use
    /slides-polish
    directly —
    /paper-talk
    would needlessly rebuild.
  • The talk content is unrelated to a paper (general lecture, demo). The orchestration assumes a paper-grounded talk; for ad-hoc decks, use
    /paper-slides
    directly with manual outline.
  • 论文尚未编译或论点未稳定。请先运行
    /paper-writing
  • 用户需要海报而非演讲。请使用
    /paper-poster
  • 用户已有完成的幻灯片,仅需视觉优化。请直接使用
    /slides-polish
    ——
    /paper-talk
    会不必要地重新构建幻灯片。
  • 演讲内容与论文无关(通用讲座、演示)。该流程假设演讲基于论文;对于临时幻灯片,请直接使用
    /paper-slides
    并手动提供大纲。

Prior Skill Relationship

与其他技能的关联

  • Composes
    /paper-slides
    ,
    /slides-polish
    ,
    /paper-claim-audit
    ,
    /citation-audit
    .
  • Does not call
    /kill-argument
    by default — that is upstream intellectual stress-test, not deck QA. Users who want talk-story stress-test before slide build can run
    /kill-argument paper/
    first.
  • Sister workflow to
    /paper-writing
    (paper) and
    /paper-poster
    (poster).
  • 整合了
    /paper-slides
    ,
    /slides-polish
    ,
    /paper-claim-audit
    ,
    /citation-audit
  • 默认不会调用
    /kill-argument
    ——该技能属于上游观点压力测试,而非幻灯片质量保证。若用户希望在构建幻灯片前对演讲内容进行压力测试,可先运行
    /kill-argument paper/
  • /paper-writing
    (论文)和
    /paper-poster
    (海报)为姐妹工作流。

Empirical Origin

经验来源

This workflow generalises a 30+ iteration polish run on a Chinese-spoken academic conference talk (May 2026). The convergent learning was that talk preparation has the same shape as paper preparation — plan, build, polish, audit, export, report — and benefits from the same assurance ladder. The per-page Codex polish pass (Phase 3) is the single most expensive but highest-value step, and is hidden behind the
polished
default so users get it without thinking about it.
该工作流基于2026年5月一场中文学术会议演讲的30+次迭代优化总结而来。核心发现是演讲准备与论文准备的流程一致——规划、构建、优化、审计、导出、报告——并可从相同的合规等级阶梯中获益。逐页Codex优化步骤(阶段3)是成本最高但价值最大的环节,因此默认启用
polished
等级,让用户无需额外设置即可获得该功能。",