fish-tts
by gtank1
DevOps & CloudBrief
This skill uses **Fish Audio S1** to generate high-quality text-to-speech audio and upload it to NextCloud.
Description
What This Skill Does
Converts text to speech using a self-hosted Fish Audio S1 instance and uploads the resulting MP3 to NextCloud via WebDAV. Supports multiple voices across languages and emotion tags for prosody control.
Combines TTS generation and cloud upload into a single scripted step, removing the need to manually transfer audio files after generation.
When to Use It
- Narrating daily notes and saving them to cloud storage
- Generating voiceovers for video scripts with a specific voice style
- Batch-producing audio files organized by date in NextCloud
- Adding emotion-tagged speech to automated notification workflows
- Testing different voices before committing to a long-form audio project
Example Workflow
User asks: generate an excited-tone greeting with af_sarah and upload it to NextCloud
- 1
Select voice af_sarah and add [excited] emotion tag to input text
- 2
POST request to Fish Audio S1 API with model, text, and voice parameters
- 3
Save response audio to a temp MP3 file
- 4
PUT the MP3 to NextCloud via WebDAV using NEXTCLOUD_USER and NEXTCLOUD_PASS
MP3 file stored at /Openclaw/ on the NextCloud instance, timestamped filename confirmed
Requirements
- Fish Audio S1 service running and accessible (FISH_AUDIO_S1_URL environment variable)
- NEXTCLOUD_USER environment variable
- NEXTCLOUD_PASS environment variable
- NEXTCLOUD_URL environment variable
- NextCloud instance with WebDAV enabled
Installation
clawhub install gtank1/fish-ttsRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.