ai-podcast — Claude Code skill: what it does, how to install
ai-podcast builds multi-character talking head video podcasts from scratch using inference.sh tools for portrait rendering, speech synthesis, and clip stitching.
no license 238 installs ★ 31 updated 2026-08-31
npx skills add 101-skills/superpowers --skill ai-podcast -a claude-codeWhat ai-podcast does
This skill guides the agent through an automated multi-speaker video workflow. It starts by setting up character portraits via pruna/p-image, Phota, or Gemini image models. Once dialog lines and pacing are settled, speech tracks are rendered with inworld/text-to-speech-2 at a default conversational tempo of 1.05.
Individual speaking turns are animated into video clips using pruna/p-video-avatar. Each clip must run sequentially because concurrent rendering runs out of GPU memory and triggers CUDA OOM errors. Finally, infsh/media-merger stitches the clips into an MP4 file at 24 frames per second.
You need the infsh CLI and access to the corresponding model endpoints. Preserving consistent human faces across camera angles also requires training a Phota profile beforehand with 30 to 50 reference portraits.
What it helps with
- Produce multi-character talking head podcasts without video equipment
- Synthesize realistic conversational speech turn by turn
- Animate still character portraits to match generated voice tracks
- Stitch alternating speaker angles into a finished MP4 file
Example prompts for the agent
- Create a 30-second AI podcast debating product pricing with two speakers
- Render character portraits and voice tracks for an illustrated mascot dialogue
- Animate speaker turns and stitch the final video using the inference pipeline
How to install ai-podcast
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill ai-podcast -a claude-code2Manually: put the files in place
~/.claude/skills/ai-podcast/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/ai-podcast/
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill ai-podcast -a codex2Manually: put the files in place
.agents/skills/ai-podcast/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill ai-podcast -a cursor2Manually: put the files in place
.cursor/skills/ai-podcast/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install 101-skills/superpowers guides/content/ai-podcast2Manually: put the files in place
.github/skills/ai-podcast/SKILL.md1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill ai-podcast -a gemini-cli2Platform CLI
gemini skills install https://github.com/101-skills/superpowers --path guides/content/ai-podcastManually: put the files in place
.devin/skills/ai-podcast/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill ai-podcast -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about ai-podcast
How to install ai-podcast in Claude Code?
Run npx skills add 101-skills/superpowers --skill ai-podcast -a claude-code in your terminal.
What does ai-podcast do?
It orchestrates character image generation, voice synthesis, avatar animation, and video assembly into a unified multi-speaker video podcast.
Why does avatar video generation need to run sequentially?
Running multiple pruna/p-video-avatar jobs concurrently exhausts GPU VRAM, causing CUDA out-of-memory failures.