一键在 Manus 中运行任何 Skill

create-talk

星标1

分支0

更新时间2026年3月3日 14:10

Generate academic presentations (Beamer .tex or Quarto RevealJS .qmd) by dispatching the storyteller agent (creator) and discussant agent (critic). Supports 4 formats — job market, seminar, short, lightning. Derives all content from paper/main.tex.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

jfimbett

jfimbett/cloco

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Create Talk

Generate an academic presentation by dispatching the storyteller (creator) and discussant (critic).

Input: $ARGUMENTS — format name, output type (beamer or quarto), optionally followed by paper path.

Workflow

Step 1: Parse Arguments

Format (required): job-market | seminar | short | lightning
Output type (optional): beamer (default) | quarto
Paper path (optional): defaults to paper/main.tex
If no format specified, ask the user before proceeding.

Format Constraints

Format	Slides	Duration	Content Scope
Job market	40-50	45-60 min	Full story, all results, mechanism, robustness
Seminar	25-35	30-45 min	Motivation, main result, 2 robustness, conclusion
Short	10-15	15 min	Question, method, key result, implication
Lightning	3-5	5 min	Hook, one result, so-what

Output Format Details

Beamer (.tex)

Output: talks/beamer/[format]_talk.tex
Compile: latexmk -pdf -cd talks/beamer/[format]_talk.tex
Figures: \includegraphics[width=0.85\textwidth]{../../paper/figures/fig.pdf}

Quarto RevealJS (.qmd)

Output: talks/quarto/[format]_talk.qmd
Compile: quarto render talks/quarto/[format]_talk.qmd
Figures: ![](../../paper/figures/fig.pdf){width=85%}
YAML: format: revealjs, theme: simple, embed-resources: true, slide-number: true

Step 2: Launch storyteller Agent

Delegate to the storyteller agent via Task tool:

Prompt: Create a [format] talk from [paper] in [beamer|quarto] format.
Read paper/main.tex fully. Extract: research question, identification strategy,
main result (with exact value and units), secondary results, robustness checks,
key figures/tables (with file paths), institutional background.
Design narrative arc for [format] format.
Build [Beamer .tex | Quarto .qmd] file. Use figures from paper/figures/ directly.
Compile to verify it runs cleanly.
Save to talks/[beamer|quarto]/[format]_talk.[tex|qmd]

The storyteller follows these principles:

One idea per slide
Figures > tables (move tables to backup slides)
Build tension: motivation → question → method → findings → implications
All claims must appear verbatim in paper/main.tex (single source of truth)
Notation must match paper exactly (same subscripts, Greek letters, variable names)

Step 3: Launch discussant Agent (Talk Critic)

After storyteller returns, delegate to the discussant agent:

Prompt: Review the talk at talks/[format]/[format]_talk.[tex|qmd].
Also read paper/main.tex to verify content fidelity.
Format: [Beamer | Quarto]
Type: [format]
Run all 8 audit categories. Score as advisory (non-blocking).
Save report to quality_reports/[format]_talk_review.md

Step 4: Fix Critical Issues

If discussant finds Critical issues (compilation failure, content not in paper):

Re-dispatch storyteller with the specific issues listed (max 3 rounds per three-strikes.md)
Re-run discussant to verify

Step 5: Present Results

Generated file path and format
Slide count and format compliance
discussant score (advisory, non-blocking)
Any TODO items (missing figures, unresolved references)

Output

Beamer: talks/beamer/[format]_talk.tex
Quarto: talks/quarto/[format]_talk.qmd

Principles

Paper is authoritative. Every claim must appear in paper/main.tex.
Same figures. Never redraw figures — use the exact same files from paper/figures/.
Less is more. Especially for short and lightning — ruthlessly cut.
Audience calibration. Job market = rigor. Seminar = interesting result. Lightning = sell the idea.
Advisory scoring. Talk scores don't block commits.
Worker-critic pairing. storyteller creates, discussant critiques. Never skip the review.

同仓库更多 Skills

同仓库

migrate-overleaf

jfimbett/cloco

Sync the overleaf/ folder with the current content of paper/. Copies all LaTeX source, figures, tables, appendix, and bibliography. Puts the model section as overleaf/model.tex (top-level) for easy coauthor access. Run whenever paper/ changes and you want to push to Overleaf.

2026-06-171

onboard-ra

jfimbett/cloco

RA onboarding audit — explore all project folders and code, inspect data files directly, and produce a structured project-state document that enables a clean restart. Use at the start of a project hand-off or when picking up work from multiple prior contributors.

2026-06-171

verify-citations

jfimbett/cloco

Full bibliography audit for an academic paper. Verifies every citation is (1) structurally correct in the .bib file, (2) actually cited in the paper, and (3) a real publication — author names, year, journal, and title verified via web search. Flags hallucinated or inaccurate references.

2026-06-171

identify

jfimbett/cloco

Design identification strategy by dispatching the causal-strategist agent (proposer) and identification-critic agent (critic). Produces strategy memo, pseudo-code, robustness plan, and falsification tests. Use when asked to "identify the effect", "design the strategy", or "write the strategy memo".

2026-03-041

context-status

jfimbett/cloco

Show current context status and session health. Use to check how much context has been used, whether auto-compact is approaching, and what state will be preserved.

2026-03-041

pipeline-status

jfimbett/cloco

Show the current research pipeline status for the active project, or display the pipeline guide for a specific project type. Reads the research spec, research journal, and active plans to determine phase, scores, and next steps.

2026-03-041

name	create-talk
description	Generate academic presentations (Beamer .tex or Quarto RevealJS .qmd) by dispatching the storyteller agent (creator) and discussant agent (critic). Supports 4 formats — job market, seminar, short, lightning. Derives all content from paper/main.tex.
disable-model-invocation	true
argument-hint	[format: job-market \| seminar \| short \| lightning] [output: beamer \| quarto (default: beamer)] [paper path (optional)]
allowed-tools	["Read","Grep","Glob","Write","Edit","Task","Bash"]

Create Talk

Generate an academic presentation by dispatching the storyteller (creator) and discussant (critic).

Input: $ARGUMENTS — format name, output type (beamer or quarto), optionally followed by paper path.

Workflow

Step 1: Parse Arguments

Format (required): job-market | seminar | short | lightning
Output type (optional): beamer (default) | quarto
Paper path (optional): defaults to paper/main.tex
If no format specified, ask the user before proceeding.

Format Constraints

Format	Slides	Duration	Content Scope
Job market	40-50	45-60 min	Full story, all results, mechanism, robustness
Seminar	25-35	30-45 min	Motivation, main result, 2 robustness, conclusion
Short	10-15	15 min	Question, method, key result, implication
Lightning	3-5	5 min	Hook, one result, so-what

Output Format Details

Beamer (.tex)

Output: talks/beamer/[format]_talk.tex
Compile: latexmk -pdf -cd talks/beamer/[format]_talk.tex
Figures: \includegraphics[width=0.85\textwidth]{../../paper/figures/fig.pdf}

Quarto RevealJS (.qmd)

Output: talks/quarto/[format]_talk.qmd
Compile: quarto render talks/quarto/[format]_talk.qmd
Figures: ![](../../paper/figures/fig.pdf){width=85%}
YAML: format: revealjs, theme: simple, embed-resources: true, slide-number: true

Step 2: Launch storyteller Agent

Delegate to the storyteller agent via Task tool:

Prompt: Create a [format] talk from [paper] in [beamer|quarto] format.
Read paper/main.tex fully. Extract: research question, identification strategy,
main result (with exact value and units), secondary results, robustness checks,
key figures/tables (with file paths), institutional background.
Design narrative arc for [format] format.
Build [Beamer .tex | Quarto .qmd] file. Use figures from paper/figures/ directly.
Compile to verify it runs cleanly.
Save to talks/[beamer|quarto]/[format]_talk.[tex|qmd]

The storyteller follows these principles:

One idea per slide
Figures > tables (move tables to backup slides)
Build tension: motivation → question → method → findings → implications
All claims must appear verbatim in paper/main.tex (single source of truth)
Notation must match paper exactly (same subscripts, Greek letters, variable names)

Step 3: Launch discussant Agent (Talk Critic)

After storyteller returns, delegate to the discussant agent:

Prompt: Review the talk at talks/[format]/[format]_talk.[tex|qmd].
Also read paper/main.tex to verify content fidelity.
Format: [Beamer | Quarto]
Type: [format]
Run all 8 audit categories. Score as advisory (non-blocking).
Save report to quality_reports/[format]_talk_review.md

Step 4: Fix Critical Issues

If discussant finds Critical issues (compilation failure, content not in paper):

Re-dispatch storyteller with the specific issues listed (max 3 rounds per three-strikes.md)
Re-run discussant to verify

Step 5: Present Results

Generated file path and format
Slide count and format compliance
discussant score (advisory, non-blocking)
Any TODO items (missing figures, unresolved references)

Output

Beamer: talks/beamer/[format]_talk.tex
Quarto: talks/quarto/[format]_talk.qmd

Principles

Paper is authoritative. Every claim must appear in paper/main.tex.
Same figures. Never redraw figures — use the exact same files from paper/figures/.
Less is more. Especially for short and lightning — ruthlessly cut.
Audience calibration. Job market = rigor. Seminar = interesting result. Lightning = sell the idea.
Advisory scoring. Talk scores don't block commits.
Worker-critic pairing. storyteller creates, discussant critiques. Never skip the review.