Deepdub TTS

Generate speech audio using Deepdub and attach it as a MEDIA file (Telegram-compatible).

Security Scan

The skill's code, declared requirements, and runtime instructions are consistent with its stated purpose (calling Deepdub to produce an MP3 and returning a MEDIA: path).

Purpose & Capability

Name/description, required env vars (DEEPDUB_API_KEY, DEEPDUB_VOICE_PROMPT_ID), required binary (python3), requirements.txt, SKILL.md pip install deepdub, and deepdub_tts.py all align: the skill needs the Deepdub SDK and an API key to synthesize speech.

Instruction Scope

SKILL.md instructs running the bundled deepdub_tts.py which only reads the declared env vars and writes output to OPENCLAW_MEDIA_DIR. The script prints the expected MEDIA: path for OpenClaw. It does not read unrelated files or environment variables.

Install Mechanism

There is no packaged installer (instruction-only). The docs instruct pip install deepdub (or an alternative 'uv pip install'), which is expected but means a third-party Python package will be downloaded/executed at install time — normal for this use but worth auditing or isolating before installing globally.

Credentials

The skill only requests the Deepdub API key and a voice prompt ID (plus optional locale/model and output dir) which are proportionate. A cautionary note: SKILL.md includes a public rate-limited trial API key — do not use it in production. Also avoid sending sensitive/personal data to the Deepdub API.

Persistence & Privilege

The skill does not request permanent 'always' inclusion, does not modify other skills or system configs, and only writes audio files to the specified media dir. Autonomous invocation of skills is the platform default and is not unusual here.

Assessment

This skill appears to do what it says: it uses the Deepdub Python SDK to synthesize text into an MP3 and returns a MEDIA: path. Before installing: (1) obtain and set your own DEEPDUB_API_KEY (do not rely on the trial key included in the docs), (2) avoid passing sensitive or private text to the TTS API because text is sent to Deepdub, (3) install the deepdub package in an isolated environment (venv/container) or audit it first since pip installs execute third-party code, and (4) if you do not want the agent to call this skill autonomously, disable model invocation or restrict invocation in your OpenClaw settings. Overall the skill is internally consistent with low risk if you follow these precautions.

What this skill does

This skill converts text into speech using Deepdub and returns an audio file as a MEDIA: attachment that OpenClaw can send to channels like Telegram.

Requirements

• Python 3.9+
• Deepdub API access

Permissions

This skill requires permission to:

• Execute deepdub_tts.py (the bundled script)
• Write audio files to OPENCLAW_MEDIA_DIR only (output path cannot be overridden via CLI arguments)

Setup

Set the following environment variables where OpenClaw runs:

Required:

• DEEPDUB_API_KEY – your Deepdub API key
• DEEPDUB_VOICE_PROMPT_ID – default voice prompt to use

Optional:

• DEEPDUB_LOCALE (default: en-US)
• DEEPDUB_MODEL
• OPENCLAW_MEDIA_DIR (default: /tmp/openclaw_media)

Free Trial Credentials

For testing only, you can use these free trial credentials:

DEEPDUB_API_KEY=dd-00000000000000000000000065c9cbfe
DEEPDUB_VOICE_PROMPT_ID=11f3403d-35b9-4817-8d55-f41694ea6227

Note: These are rate-limited trial credentials for evaluation purposes only. Do not use for production. Obtain your own API key and voice prompts from Deepdub for production use.

Install dependency

Install the official Deepdub Python SDK:

pip install deepdub

Or using uv (faster alternative):

uv pip install deepdub