wowerpoint
Turn one document into a kawaii NotebookLM slide-deck PDF. Use for "wowerpoint this", "make a deck about <file>", "turn this report into slides", or any request to render a single document as shareable narrative slides.
Security Assessment
Detected risks:
About wowerpoint
Wowerpoint turns a single document into a kawaii NotebookLM slide-deck PDF. Its scope is deliberately narrow—one doc in, one PDF out—and it is explicitly slide-deck only, noting that videos and podcasts from the same engine are noticeably worse and out of scope, with users referred to the notebooklm CLI directly for those. It triggers on requests like "wowerpoint this," "make a deck about <file>," "turn this report into slides," or "kawaii-deck this."
The workflow depends on the notebooklm-py tool (installed via uv tool install with Playwright and Chromium) and jq for JSON parsing, with a one-time setup that can be skipped if notebooklm auth check returns 0 and jq resolves. Authentication is interactive and must be completed by the user via OAuth—it cannot be scripted. The skill requires exactly one source document; if the source is missing or too thin to carry a deck, it should be written first as comprehensive long-form narrative rather than padded with additional sources. After an auth pre-flight, it creates a notebook and adds the source, carefully parsing the differing JSON envelope keys (create returns .notebook.id, source add returns .source.id, generate returns .task_id), since a wrong key yields an empty string and silent downstream failure.
Because generation takes roughly ten minutes, the skill never blocks on it—it spawns a subagent in the background (a copy-paste template is provided) that waits for the source, generates the slide deck, waits for the artifact, downloads it, and verifies the file. The output PDF is written adjacent to the source with a parallel filename (or under reports/ if no sensible location exists). After the PDF lands, the subagent optionally POSTs it to the WOWerpoint Server, which converts the 16:9 deck into a 9:16 mobile twin and returns a share URL that becomes the primary deliverable (the PDF is the backup); this step is skipped if the required environment variables are unset. The doc includes a default one-sentence kawaii prompt, share-link handling that guards against null ids, and a catalog of failure modes such as missing pip, uninstalled Playwright, un-completed OAuth login, and rate-limit errors.
FAQ
What formats can this skill produce?
Slide-deck PDFs only. Videos and podcasts from the same engine are noticeably worse and out of scope; the doc refers users to the notebooklm CLI directly if they want those.
What setup and tools are required?
It needs notebooklm-py installed via uv tool install (with Playwright and Chromium) plus jq for JSON parsing. Setup can be skipped if notebooklm auth check returns 0 and jq resolves.
How does authentication work?
Authentication is interactive OAuth that only the user can complete—it cannot be scripted. The user is told to run notebooklm login so the OAuth ENTER lands in their terminal.
How many source documents does it use, and what if the source is thin?
Exactly one source document. If it doesn't exist or is too thin to carry a deck, the skill writes it first as comprehensive long-form narrative rather than adding more sources.
What is the primary deliverable?
If the WOWerpoint Server environment variables are set, the primary deliverable is a mobile share URL (a 9:16 twin of the 16:9 deck), with the PDF on disk as the backup. If those vars are missing, the PDF is handed over directly.
Install wowerpoint
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
thedotmack/claude-mem