VoiceCheap

Use when the user asks to connect OpenClaw to their VoiceCheap account or use VoiceCheap for transcription, YouTube transcripts and chapters, subtitles, speech and permitted voice creation, images and thumbnails, dubbing, lip sync, publishing or workspace actions.

VoiceCheap for OpenClaw

Follow the workflow reference for account access, media actions, credits, retries, asynchronous status and privacy. Discover the actual catalog before choosing a transcription, subtitle, voice, image, project or workspace action; tools depend on the permissions the user grants.

If VoiceCheap is not connected, inspect existing MCP configuration with openclaw mcp status --verbose. With the user's request to connect, add the remote endpoint using the installed CLI's supported syntax:

sh
openclaw mcp add voicecheap --url https://api.voicecheap.ai/mcp --transport streamable-http
openclaw mcp login voicecheap

Preserve an existing server and other connections. Let the user complete the browser login and approve only account:read for a basic connection check. For a requested media workflow, choose only its required permissions; a Read Only preset may also expose project and media data. Do not extract or request credentials. If the CLI or browser flow is unavailable, explain the exact step the user needs to complete rather than claiming the server is configured.

Check openclaw mcp doctor voicecheap --probe, then use actual account/project read tools through the active runtime. A saved configuration or successful probe alone does not prove that an authenticated customer action succeeded. Reconnect with additional user-approved permissions only when a requested workflow needs them. No blanket tool auto-approval is required.

Keep account identifiers, private media, transcripts and credentials out of public listings and demonstrations. Use synthetic review content and the minimum permissions needed. Publishing outputs and cloning voices require specific authorization; voice samples also require the speaker's permission.