Getting Started

Prerequisites, installation, configuration, and recommended prompts for ListenHub Skills.

Prerequisites

ListenHub API Key — available to all users at API Key Settings
Dependencies — curl and jq, installed automatically by the AI
Any client with Agent Skills — Claude Code, Cursor, Windsurf, OpenCode, and others

npx skills add marswaveai/skills

bunx add-skill marswaveai/skills

ListenHub API Key — Get one here, available to all users
Image generation — uses the same API key; default output directory is ~/Downloads
On first use, the AI guides setup automatically.

Goal	Recommended Phrasing
Podcast	"Make a podcast, topic is XXX, in Chinese, be direct"
Podcast	"Make a podcast, topic is XXX, in English, go deep"
Explainer Video	"Make an explainer video, content is XXX, in Chinese, explain clearly"
Explainer Video	"Make an explainer video, content is XXX, in English, more narrative"
Voice Reading	"Turn this text into audio, in Chinese, read it directly"
Voice Reading	"Turn this link into audio, in English, do not edit the original text"
Image	"Generate image: XXX, 16:9, 2K"

If language is not specified, the system detects Chinese or English from the input.
If only a topic is provided, the podcast defaults to quick + solo.
If voice is not specified, the first available voice for the language is selected.
Voice reading uses FlowSpeech direct by default.
Multi-character dialogue in scripts uses Speech as an advanced feature.