Productivity & Collaboration ppt-generationimage-generationvideo-transitionsgemini-apikling-aiffmpegslide-deck

NanoBanana PPT Generator

Turn documents into high-quality image-based PPT slides with AI, complete with transition videos and an interactive player.

FollowSkills review · FSRS-1.0
Recommended
60/ 100 5-point scale 3.0 / 5
This review was completed under FSRS 1.0. Dimension scores are not converted; it is queued for re-review under FSRS 2.0.
1 2 3 4 5 6
1Utility14 / 20 · 3.5/5
2Reliability11 / 20 · 2.8/5
3Safety18 / 25 · 3.6/5

Requires the user's own GEMINI_API_KEY and optional Kling AI keys; generates local image/video files with no destructive operations observed — calling paid third-party APIs will incur cost.

4Evidence5 / 15 · 1.7/5
5Usability7 / 10 · 3.5/5
6Maintenance5 / 10 · 2.5/5
Evidence confidence:Low Reviewed Jul 17, 2026
Before you use it
  • Depends on paid Gemini (required) and Kling AI (optional) APIs — actual usage incurs API costs
Review evidence [1]
See the full review method →

What does this skill do, and when should you use it?

This is an open-source Claude Code Skill by 歸藏 (op7418) that uses Google Nano Banana Pro (Gemini 3 Pro Image Preview) to turn document content into 16:9 PPT images at 2K/4K resolution, with two built-in professional styles: gradient glass and vector illustration. Version 2.0 adds Kling AI-powered page transition videos, a looping cover preview animation, and an interactive HTML player, then composes everything into a full presentation video with FFmpeg. Claude handles document analysis, slide planning, and transition prompt writing, while Python scripts do the actual image and video generation. It suits users who frequently need quick presentations and are comfortable supplying their own API keys.

Reads a user-supplied document or pasted text, plans slide content for a chosen page count (5–25) and writes slides_plan.; calls the Nano Banana Pro API to generate each 16:9 slide at 2K/4K (~30–60 s per slide); Claude compares adjacent slides to write tailored transition prompts, then calls Kling AI to render 5-second 1080p/24fps transition videos; outputs an interactive HTML player (keyboard navigation) and a full video (full_ppt_video.mp4) composed with FFmpeg (H.264, CRF 23).

  1. A product manager who needs to turn meeting minutes into a 5-slide deck by pasting content directly into Claude Code
  2. Presenters preparing tech launches or business reports who want Apple-Keynote-style visuals via the gradient glass style
  3. Educators building warm, illustration-style courseware with the vector illustration style
  4. Content creators who want to share their presentation as an MP4 with transitions
  5. Scenarios needing 4K slides for print or large-screen display
  6. Python-savvy users running the scripts standalone for batch slide generation

What are this skill's strengths and limitations?

Pros
  • One pipeline from document analysis to images and video; fast at ~30 s/slide for 2K
  • Two polished built-in visual styles, extensible via the styles/ directory
  • Transition prompts are tailored by Claude against the actual generated images, with attention to text stability
  • Delivers both an interactive player and a full video, plus PDF export via print
  • Clean .env-based key management and thorough docs on install, security, and API management
Limitations
  • Depends on paid APIs: Google Gemini is required, Kling AI for video; cost scales with page count
  • The GitHub repo lists no license (README text states MIT); verify before adopting
  • Slides are large raster images (~2.5MB at 2K, ~8MB at 4K) with non-editable text
  • The slash command and AskUserQuestion flow assume Claude Code; other platforms need rework
  • No test suite in the repo; output quality depends on how structured the input document is, and each transition video takes 30–60 s to generate

How do you install this skill?

  1. git clone https://github.com/op7418/NanoBanana-PPT-Skills.git
  2. Create a virtualenv: python3 -m venv venv && source venv/bin/activate
  3. pip install google-genai pillow python-dotenv
  4. cp .env.example .env and fill in GEMINI_API_KEY (required) plus Kling AI's KLING_ACCESS_KEY / KLING_SECRET_KEY (video only)
  5. To install as a Skill: clone into ~/.claude/skills/ppt-generator/ (or run bash install_as_skill.sh)
  6. Verify: python3 generate_ppt.py --help

Video features also require FFmpeg (macOS: brew install ffmpeg; Ubuntu: sudo apt-get install ffmpeg).

How do you use this skill?

In Claude Code, run /ppt-generator-pro or simply say "Generate a 5-page PPT from this document using the gradient glass style." Claude asks for style, page count, resolution (2K/4K), and whether to generate video, then plans content and returns the output directory. Standalone mode: python3 generate_ppt.py --plan my_slides_plan. --style styles/gradient-glass.md --resolution 2K; for video, run generate_ppt_video.py with the transition_prompts. file. Open index.html in a browser to present with arrow keys, spacebar, ESC, etc.; print to PDF for export.

Related skills