Summary

WRIT-FM offers a framework for a 24/7 AI-powered internet radio station, automating music generation, DJ scripting, and text-to-speech delivery. It targets technically savvy users interested in autonomous media creation, providing a complete AI-driven content pipeline.

How It Works

The system uses a writ CLI to manage components: Claude CLI writes DJ scripts, Kokoro TTS voices them, and ACE-Step via music-gen.server generates music. Five AI hosts rotate across eight shows with distinct voices and topics. An operator_daemon.sh ensures continuous operation by health-checking and stocking content. Multiple isolated station instances are supported.

Quick Start & Requirements

Setup requires Python 3.11+, uv, and system dependencies (icecast, ffmpeg, ezstream, vorbis-tools on macOS). Kokoro TTS needs uv pip install kokoro soundfile (~200MB model). Configuration involves editing mac/config.yaml and config/icecast.xml. Start with ./writ start. Optional dependencies: Claude CLI/Codex CLI for agents, music-gen.server for AI music, cloudflared for tunneling. Apple Silicon recommended.

Highlighted Details

• Five AI hosts with unique personas, voices, and topic specializations.
• AI-generated music as primary broadcast content.
• Automated content includes hosted breaks scripted by Claude CLI and spoken by Kokoro TTS.
• Supports multiple, isolated station instances (e.g., WRIT-FM, KLOD-FM, CDEX-FM).
• Options for relaying streams to YouTube RTMP.

Maintenance & Community

The README provides no details on maintainers, community channels (Discord/Slack), sponsorships, or roadmap links.

Licensing & Compatibility

Released under the MIT license, this permissive license allows commercial use and integration into closed-source projects.

Limitations & Caveats

Full functionality, especially AI music generation and advanced agent features, depends on optional external dependencies like music-gen.server and Claude CLI/Codex CLI. Setup instructions are primarily macOS-focused with Apple Silicon recommended, potentially indicating OS-specific complexities or performance considerations.