book-to-podcast

What problem does it solve?

Producing an audiobook-style podcast from a non-fiction PDF or EPUB is slow and labor-intensive because it requires extracting text, writing a structured script, generating narration audio, and publishing an RSS feed users can subscribe to.

Core Features & Use Cases

• End-to-end book → podcast pipeline: Extracts text from EPUB/PDF, generates a 45-minute script in either monologue or two-host conversation format, renders MP3 via a TTS provider, and updates a private RSS feed stored on Cloudflare R2.
• Multiple TTS providers with Indian-language support: Supports ElevenLabs and Sarvam, plus optional fully local Kokoro; Sarvam enables multiple Indian languages using compatible models and language codes.
• Resumable, cost-aware rendering: Uses per-chunk persistent audio caching so reruns continue without re-spending credits on already rendered segments.
• Use case: Convert a newly downloaded non-fiction book into a commute-ready podcast episode, then subscribe from any podcast app via the skill’s generated R2 feed URL.

Quick Start

Run the skill on your book file by telling: "/book-to-podcast path/to/some-book.epub".