Text to Voice Local
v1.4.0Local text-to-voice generation for OpenClaw workspaces using a canonical txt-to-mp3 pipeline. Use when the user wants to turn any prepared text into voice, a...
⭐ 1· 90·0 current·0 all-time
byIldar@betonimig
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (local text-to-voice) align with the provided scripts: they read text files, call a Node TTS helper, use ffmpeg to concat/encode chunks, and update canonical output/state files. Requested binaries (node, ffmpeg, node-edge-tts) are expected for this functionality.
Instruction Scope
SKILL.md and the shell/js scripts limit their actions to workspace tmp and skill state paths; they read text files, produce mp3s, and update a last-output pointer. There are no instructions to read unrelated system files, access unrelated credentials, or transmit data to custom external endpoints. Note: the node dependency node-edge-tts will perform TTS and likely contact the remote TTS service as part of normal operation (this is expected for a TTS library).
Install Mechanism
No automated install spec is present (the skill is instruction + scripts). The README tells the operator how to install node-edge-tts (npm i -g). That is reasonable, but users should be aware this requires running a global npm install and network access to npm; the skill itself does not embed or download external binaries automatically.
Credentials
The skill does not request environment variables or external credentials. It uses OPENCLAW_WORKSPACE optionally and TTS_RETRIES/other shell variables if present — all proportional to operation. The included state files contain absolute example paths under /root/.openclaw/workspace, which are benign but operator-specific and should be adjusted when installed into another workspace.
Persistence & Privilege
always:false and user-invocable; scripts only write to their own skill state directory and workspace tmp. The skill does not modify other skills or global agent settings.
Assessment
This skill appears to do what it says: convert local text files to MP3 using node + node-edge-tts and ffmpeg, and maintain a canonical latest.mp3 path. Before installing/using it: (1) review and decide whether you’re comfortable installing node-edge-tts globally (npm i -g), since that requires network access to npm and the TTS library will contact its remote TTS provider during generation; (2) run the scripts in a non-root workspace or inspect/adjust the state JSON canonical paths (the bundled state uses /root/.openclaw/...); (3) verify node and ffmpeg come from trusted sources for your environment; (4) if you need offline-only operation, confirm node-edge-tts’s privacy/endpoint behavior or replace it with an explicitly offline engine. Otherwise the skill is coherent and proportionate to its stated purpose.Like a lobster shell, security has layers — review code before you run it.
latestvk97c61fb38edtrkt0ctekxbq1x84bej7
License
MIT-0
Free to use, modify, and redistribute. No attribution required.