Skip to content

2654400439/Paper2Seminar / skills/paper-ppt-orchestrator

paper-ppt-orchestrator

Verified

Build complete fixed-layout academic-paper presentations from a PDF with bundled PPTX and one-page HTML templates. Use for full paper reading, complexity-aware slide budgeting, subsection coverage mapping, figure extraction, TikZ or Matplotlib evidence visuals, deterministic PPTX assembly, and final QA. Trigger for requests to turn a paper PDF into a discussion-group deck, especially 12-16+ page systems, security, networking, or measurement papers.

FileKindSize
SKILL.mdskill md17.9 KB
agents/openai.yamlconfig296 B
assets/one-page-summary.htmlother7.8 KB
assets/seminar-template.pptxbinary572.7 KB
references/content-design.mdmarkdown5.7 KB
references/data-viz.schema.jsonconfig2.9 KB
references/deck-plan.schema.jsonconfig8.5 KB
references/deck-sizing.mdmarkdown3.4 KB
references/environment-compatibility.mdmarkdown2.8 KB
references/feature-registry.jsonconfig1.4 KB
references/figure-extraction.mdmarkdown5.1 KB
references/job-request.defaults.jsonconfig1.6 KB
references/job-request.schema.jsonconfig7.5 KB
references/one-page-template.mdmarkdown3.1 KB
references/qa-gates.mdmarkdown10.0 KB
references/speaker-notes.mdmarkdown3.1 KB
references/template-contract.mdmarkdown5.1 KB
references/tikz-layout.mdmarkdown12.2 KB
references/visual-selection.mdmarkdown6.5 KB
references/workflow.mdmarkdown9.8 KB
requirements-doclayout.txtmarkdown238 B
requirements-ui.txtmarkdown39 B
requirements.txtmanifest68 B
scripts/apply_speaker_notes.pyscript6.1 KB
scripts/build_ppt.pyscript16.3 KB
scripts/crop_author_block.pyscript7.9 KB
scripts/crop_image_region.pyscript2.4 KB
scripts/export_pptx_pages.ps1script1.3 KB
scripts/extract_paper_assets.pyscript22.3 KB
scripts/job_request.pyscript17.1 KB
scripts/make_contact_sheet.pyscript4.0 KB
scripts/paper_ppt.pyscript17.1 KB
scripts/preflight_capabilities.pyscript8.3 KB
scripts/render_data_viz.pyscript14.6 KB
scripts/render_deck_overview.pyscript2.4 KB
scripts/render_one_page_html.pyscript13.4 KB
scripts/render_readability.pyscript4.1 KB
scripts/update_qa_status.pyscript10.1 KB
scripts/validate_deck_plan.pyscript33.9 KB
webapp/__init__.pyscript49 B
webapp/backend.pyscript14.5 KB
webapp/job_runner.pyscript3.2 KB
webapp/streamlit_app.pyscript17.3 KB

references/deck-sizing.md

View on GitHub
# Deck sizing and coverage

## Principle

Derive slide count from paper complexity and coverage, not from the number of template content slides. Treat target duration as a pacing constraint, not permission to omit central methods or experiments.

The bundled template contributes nine non-content slides: cover, paper information, one-page table, contents, four section dividers, and closing. Count these separately from content slides.

## Default ranges

| Paper profile | Typical evidence | Total deck | Content slides |
|---|---|---:|---:|
| Short paper, 6-8 pages | One main idea and compact evaluation | 18-24 | 9-15 |
| Standard paper, 9-11 pages | One method plus normal evaluation | 22-28 | 13-19 |
| Full paper, 12-16+ two-column pages | Multi-module method or substantial measurement/evaluation | 26-32 | 17-23 |
| Measurement plus new system/defense | Two substantial contributions and multiple RQs | 28-32 | 19-23 |

For a full-length paper, do not produce fewer than 26 total slides unless the user explicitly requests a short talk. If a shorter deck is requested, record the omitted subsections and describe the result as an abbreviated deck.

## Section budgets for a full systems/security paper

Use these as starting ranges, then adapt to the paper:

- Background, threat model, benchmark, or measurement context: 4-6 content slides.
- Method: 5-7 content slides. Allocate an overview, at least one slide per substantive module, and separate slides for algorithms, optimization loops, restoration logic, or implementation details when they carry independent ideas.
- Results: 6-9 content slides. Separate experimental setup, primary effectiveness, baseline/ablation, robustness or sensitivity, generalization, usability, and overhead when present.
- Reflection: 2-3 content slides.

Do not force equal section lengths. Do not interpret the template's one seed content slide per section as a three-page or four-page cap.

## Mandatory coverage matrix

Before writing slide bodies, create a matrix in `paper-notes.json` with one row per main paper subsection:

```json
{
  "paper_section": "7.3 Robustness",
  "importance": "core",
  "planned_slide_ids": ["S24", "S25"],
  "status": "covered",
  "omission_reason": ""
}
```

Allowed status values are `covered`, `notes_only`, and `omitted`. Every abstract contribution, main method subsection, research question, primary experiment, limitation, and major appendix result used by the paper's claims must be `covered` or `notes_only`. A core item cannot be `omitted` without explicit user approval.

Run these checks before asset generation:

- Every paper contribution maps to at least two content slides when it contains multiple stages, modules, or studies.
- Every substantive method module maps to at least one slide.
- Experimental setup and metrics are visible before result interpretation.
- Every named RQ maps to at least one result slide.
- Baselines/ablations, adaptive or sensitivity tests, usability, and overhead are not silently collapsed when they support different claims.
- Reading only the title sequence still exposes the full argument and evidence chain.

## Per-slide budget

Keep the existing fixed layout:

- One idea and one takeaway per content slide.
- Normally three bullets totaling 5-6 rendered lines at 18pt.
- Split a slide instead of shrinking text or combining independent claims.
- Use speaker notes for qualifications and transitions, not for hiding an entire omitted subsection.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 196kUpdated Claude CodeCodexVerifiedInstall