Send high-quality text-to-speech voice messages on WhatsApp in 40+ languages with automatic delivery
Security Analysis
high confidenceThe skill's files and runtime instructions are internally consistent with a TTS→WhatsApp sender that uses Piper, FFmpeg, and the user's Clawdbot installation; nothing requested or instructed appears disproportionate or unrelated to its stated purpose.
Name/description match the instructions: Piper TTS -> FFmpeg conversion -> send via Clawdbot. Prerequisites (Piper, FFmpeg, Clawdbot, voice models) and behavior (OGG/Opus conversion, cleanup, group support) are appropriate for a WhatsApp TTS sender.
SKILL.md instructs the agent/user to generate TTS, convert audio, and send via Clawdbot; it references only expected files/paths (~/.clawdbot/*, models directory) and Clawdbot CLI commands. No instructions leak or collect unrelated system data or send data to unexpected external endpoints.
This is an instruction-only skill (no install spec, no code). Installation guidance relies on standard package sources (pip, system package managers) and model downloads from Hugging Face — these are expected and traceable. No downloads from obscure shorteners or personal IPs are instructed.
The metadata lists no required env vars, but the SKILL.md refers to environment/config values (PIPER_BIN_PATH, WHATSAPP_DEFAULT_TARGET, PIPER_DEFAULT_*) stored in ~/.clawdbot/clawdbot.json and depends on Clawdbot's existing WhatsApp connection/credentials. This is proportionate to the functionality but users should be aware the skill will rely on Clawdbot's WhatsApp credentials and local model files.
Skill is user-invocable, not always-on, and does not request permanent platform-wide privileges. It instructs editing its own Clawdbot skill config, which is normal. No evidence it modifies other skills or global agent settings outside its scope.
Guidance
This skill appears to do what it says: generate Piper TTS audio, convert it, and send via your installed Clawdbot (which must already be connected to WhatsApp). Before installing: 1) Confirm Clawdbot and WhatsApp are configured and you’re comfortable with those credentials being used for sends; 2) review where model files are stored (~/.clawdbot/skills/piper-tts/models/) and verify Hugging Face downloads and model licenses; 3) test with --no-send to verify audio generation and conversion locally; 4) be cautious when enabling automatic sends (use WHATSAPP_DEFAULT_TARGET or specify --target deliberately). Metadata did not declare the env vars the README references, so double-check your ~/.clawdbot/clawdbot.json configuration if you rely on defaults.
Latest Release
v1.0.0
Initial release - TTS voice messages in 40+ languages
More by @hopyky
Published by @hopyky on ClawHub
