explainer
Use when the user asks for explainer or a task matching the examples below. ~60-80s explainer video for any URL — GitHub repo, product page, docs site, blog post, or launch. Canonical workflow for URL walkthroughs. Use when the user asks to "explain this URL / repo / website / product", "make a walkthrough video for [url]", "demo this site", "Loom-style explainer of [url]", "explainer for github.com/...", or "explain this product link". Drives a real browser through the URL, generates an avatar
Security Assessment
About explainer
Explainer is a Pika Labs plugin skill that generates a roughly 60-80 second explainer/walkthrough video for any URL, whether a GitHub repo, product page, docs site, blog post, or launch. It solves the problem of quickly producing a Loom-style demo: it drives a real browser through the URL along a beat-sheet timeline, generates an avatar lipsync of the narration, and composites everything into a 1280x800 macOS Sonoma frame with a bottom-left circular presenter avatar and element-targeted zoom on each mid-section beat.
The skill detects GitHub URLs and switches into a repo-aware mode that scans the README and detects a live demo, while all other URLs use a generic page-walkthrough flow. It favors firing fast: it resolves the avatar and voice silently (honoring explicit --avatar and --voice overrides, otherwise generating a presenter portrait and default voice) and makes a confident first attempt at focus angles rather than prompting. The only mandatory confirmation is a cost-transparency gate (estimated 50-500 credits / ~$0.50-$5.00 depending on lipsync provider, captions, and preview mode); an additional preview guardrail becomes mandatory when the avatar is a generated or regenerated fallback. Options include lipsync provider choice (pika vs kling), captions on/off, a live-demo URL override, and preview mode. It also documents that inline-pasted images on Claude Desktop cannot reach MCP tools, guiding users to a hosted URL or file upload.
Target users are developers, founders, and product marketers who want an automated narrated demo of a link without recording it themselves, and who value a repo-aware mode for open-source projects.
FAQ
What kinds of URLs does it handle?
Any https URL: GitHub repos (which activate a repo-aware README scan and live-demo detection), product/launch pages, docs sites, and blog posts. Non-GitHub URLs use a generic page-walkthrough flow.
Do I have to supply an avatar and voice?
No. It resolves them silently, generating a presenter portrait and default voice if none are given, but honors explicit --avatar and --voice overrides.
What does it cost and when does it ask me?
The cost gate estimates about 50-500 credits (~$0.50-$5.00) depending on lipsync provider, narration, captions, and preview. That is the main confirmation; a preview is also forced when the avatar is a generated/regenerated fallback.
What is the difference between the pika and kling lipsync providers?
Pika (the default) is faster (~2-5 min) with slightly more head motion; kling is tighter and face-centered but slower (~5-30 min), recommended for high-stakes renders.
Why can't I paste an image directly on Claude Desktop?
Inline-pasted images don't reach MCP tools yet (an Anthropic-side limitation). Provide a hosted URL, or attach a file so it can be uploaded to a public URL first.
Install explainer
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
pika-labs/pika-plugins