← Back to Skills Marketplace
566
Downloads
1
Stars
6
Active Installs
4
Versions
Install in OpenClaw
/install noizai-tts
Description
Use this skill whenever the user wants to convert text into speech, generate audio from text, or produce voiceovers. Triggers include: any mention of 'TTS',...
Usage Guidance
This skill appears to do what it says — convert text to speech using either a local Kokoro CLI or the Noiz cloud API. Before installing, consider the following:
- Credential handling: If you configure a NOIZ_API_KEY it will be normalized and saved to ~/.config/noiz/api_key (permissions forced to 0600). If you have an old ~/.noiz_api_key it will be copied to the new path (not deleted). If you do not trust that legacy file, remove it first.
- Network activity: The skill will call Noiz endpoints (default base https://noiz.ai/v1) and may download reference audio from URLs you supply (or its default sample URLs stored on storage.googleapis.com / noiz.ai). Only provide reference_audio URLs you trust.
- Local dependencies: The timeline/rendering features use ffmpeg/ffprobe and (for the local backend) kokoro-tts; the Python 'requests' package is required for Noiz backends. The registry metadata does not list these required binaries — ensure they are available or use guest mode / Kokoro as appropriate.
- Data flow: Uploaded reference audio or generated audio is sent to Noiz when using the cloud backend (authenticated with your API key). If you want to avoid sending data to the cloud, use the Kokoro backend or avoid providing an API key.
- Review defaults: The script points to default reference audio files hosted externally; if you are concerned about unexpected downloads, replace or remove those defaults.
If any of the above is unacceptable, do not install; otherwise the skill is coherent with its stated purpose. If you want higher assurance, review the included Python files yourself and test guest mode first (no API key required) to validate behavior.
Capability Analysis
Type: OpenClaw Skill
Name: noizai-tts
Version: 1.0.3
The skill bundle provides comprehensive text-to-speech capabilities using Kokoro (local) and Noiz (cloud) backends. It manages API keys securely by storing them in `~/.config/noiz/api_key` with restricted permissions (0600) and includes a transparent 'Security & data disclosure' section in `SKILL.md` detailing its file and network operations. The code uses standard libraries and subprocess calls (e.g., `ffmpeg`, `ffprobe`) with safe argument lists, and its network activity (communicating with `noiz.ai` and downloading reference audio from user-specified URLs) is strictly aligned with its stated purpose.
Capability Assessment
Purpose & Capability
Name/description (text-to-speech, voice cloning, timeline rendering) match the included scripts and capabilities (kokoro local backend, Noiz cloud backend, timeline rendering, SRT handling, voice maps). Requesting NOIZ_API_KEY as primary credential is appropriate for the Noiz cloud features.
Instruction Scope
Runtime instructions and scripts call out network and filesystem operations (downloading reference audio, posting text/audio to Noiz endpoints, reading/writing SRT/audio files, invoking ffmpeg/kokoro-tts). This is expected for a TTS skill, but the code will download arbitrary reference_audio URLs you provide and will access ~/ .noiz_api_key (legacy) for migration; review any URLs and local paths you pass to the tool.
Install Mechanism
There is no install spec (instruction-only), which limits automatic installation risks. However, the scripts depend on external programs/libraries (ffmpeg/ffprobe, kokoro-tts CLI for local backend, and the Python 'requests' package) that are not declared as required binaries in the registry metadata. The code also writes files (temp audio, final outputs).
Credentials
The only cloud credential requested is NOIZ_API_KEY (declared as primaryEnv), which aligns with the cloud backend features. The registry listing shows 'required env vars: none' but metadata/SKILL.md do identify NOIZ_API_KEY as primary credential — this minor mismatch is informational rather than malicious.
Persistence & Privilege
The skill writes the API key to ~/.config/noiz/api_key (0600) and will copy a legacy ~/.noiz_api_key into that location if present (non-destructive copy). It also creates temporary files and output audio under user-specified paths. It does not request always:true or modify other skills; persistence is limited to its own config file.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install noizai-tts - After installation, invoke the skill by name or use
/noizai-tts - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.3
- Added explicit permissions section listing `network` and `filesystem` requirements.
- Metadata updated to indicate `NOIZ_API_KEY` as the primary environment variable.
- No functional or file changes; documentation has been clarified for permissions and environment variable usage.
v1.0.2
- Added explicit environment variable support for NOIZ_API_KEY and listed it as an optional configuration method.
- Declared runtime dependencies and outlined which are required or optional (requests, ffmpeg, kokoro-tts).
- Documented network and file access in detail, including precise endpoint URLs and credential file locations.
- Described security, credential storage, and data disclosure behavior, including key migration and temporary file handling.
- Made it clear which operations trigger network or file access, and clarified backend-specific behavior (e.g., Kokoro works fully offline).
- No code or feature changes—documentation and security transparency update only.
v1.0.1
- Added `requests` package as a requirement for Noiz backend.
- Minor correction to the SKILL.md formatting and description.
- No functional code or feature changes detected.
v1.0.0
Initial release of the tts skill.
- Converts text to speech with support for both Kokoro (local) and Noiz (cloud) backends.
- Supports simple one-off text-to-audio, timeline-accurate SRT dubbing, voice cloning, emotion, speed adjustments, and per-segment voice mapping.
- Guest mode provides limited TTS features without requiring an API key.
- Commands documented for various use cases; supports integration with Feishu, Telegram, Discord.
- Timeline mode allows for SRT to audio, per-segment configuration, dynamic reference audio, and voice maps.
- Includes table of available guest voices for quick reference.
Metadata
Frequently Asked Questions
What is tts?
Use this skill whenever the user wants to convert text into speech, generate audio from text, or produce voiceovers. Triggers include: any mention of 'TTS',... It is an AI Agent Skill for Claude Code / OpenClaw, with 566 downloads so far.
How do I install tts?
Run "/install noizai-tts" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is tts free?
Yes, tts is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does tts support?
tts is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created tts?
It is built and maintained by kusuriuri (@ksuriuri); the current version is v1.0.3.
More Skills