Claude Desktop

Tools

Coming soon. Reference for the Wubble tools submitted for Anthropic review in the Claude Desktop bundle.

Submitted for review

These are the tools in the prepared Wubble Desktop Extension. Public availability depends on Anthropic accepting or listing the extension in Claude Desktop.

Tool reference

When the extension is available, Claude Desktop can choose tools from Wubble when you ask it to check allowance, generate audio, wait for completion, inspect Wubble-owned generated audio, or load a completed audio result.

Current submitted surface

The first submitted bundle focuses on Wubble-owned generation workflows and avoids local file upload or Claude attachment ingestion tools.

Read and playback tools

Tool	Use	Scope	Prompt
`get_credits`	Read plan allowance, remaining credits, API calls, and reset timing.	`analytics:read`	Check my Wubble credits.
`get_usage`	Summarize recent Wubble usage by category with compact recent records.	`analytics:read`	Show my recent Wubble usage by category.
`get_request_status`	Read one Wubble request status. Also accepts latest, last, or most_recent.	`analytics:read`	Check the latest Wubble request.
`wait_for_request`	Wait briefly for a Wubble request and return completed audio when ready.	`analytics:read`	Wait for the latest Wubble request and load it when ready.
`list_voices`	Search and summarize available voices for TTS and dialog workflows.	`analytics:read`	Find a warm narrator voice.
`get_speech_status`	Read speech-specific status for completed or processing speech requests.	`analytics:read`	Check the latest speech request.
`load_audio`	Render a safe Wubble-hosted audio URL in the bundled audio player.	`none`	Load the audio from the last result.

Generation and editing tools

Tool	Use	Scope	Prompt
`generate_tts`	Generate text-to-speech from text and optional voice settings.	`audio:generate`	Generate a calm voiceover from this paragraph.
`generate_music`	Generate instrumental music only.	`audio:generate`	Make an upbeat product demo bed.
`generate_song`	Generate a full song from required lyrics.	`audio:generate`	Generate lyrics for a lofi chorus, then create a song from them.
`generate_lyrics`	Generate lyrics text without creating playable audio.	`audio:generate`	Write lyrics for a brand launch jingle.
`extend_lyrics`	Extend existing lyrics text.	`audio:generate`	Extend these lyrics with a second verse.
`extend_song`	Extend a completed Wubble song by song ID.	`audio:generate`	Extend the latest completed Wubble song.
`generate_track`	Create a new track from a completed Wubble song ID and instrument group.	`audio:generate`	Generate a bass track for this completed Wubble song.
`edit_track`	Edit a completed Wubble song or instrumental track request.	`audio:generate`	Edit the latest instrumental track to make the intro brighter.
`generate_sound_effect`	Generate a short sound effect from a prompt.	`audio:generate`	Create a soft notification sound.
`text_to_dialog`	Generate dialog audio from multiple speaker turns.	`audio:generate`	Create a two-speaker welcome dialog.
`describe_audio`	Describe Wubble-owned generated audio from a URL or completed request.	`audio:generate`	Describe the latest generated audio.
`separate_stems`	Create stem package links from Wubble-owned generated audio.	`audio:generate`	Separate stems from the latest completed song.
`remix_song`	Remix a Wubble song from song ID and lyrics.	`audio:generate`	Remix this Wubble song with these new lyrics.

Generation consumes allowance

Tools with audio:generate can consume Wubble requests or credits. Claude Desktop will ask before running tools that the extension marks as spendful.

API key scopes

analytics:read

Required for allowance, usage, voice listing, request status, and completed result checks.

audio:generate

Required for TTS, music, songs, lyrics, tracks, sound effects, dialog, audio analysis, stem separation, and remixing.

Generated audio

Most longer audio workflows return quickly with a Wubble request. Claude can use wait_for_request with latest, then call load_audio when a playable Wubble URL is ready.

GET/v1/requests/:requestId

Auth: API keyBilling: Free status readBehavior: Status and completion pollingScopes: analytics:read

Example promptstext

Check my Wubble credits.
Find a warm narrator voice.
Generate a calm voiceover using a warm narrator style.
Generate lyrics for a launch jingle, then create a song from those lyrics.
Create a short glass notification sound.
Describe the latest generated audio.
Separate stems from the latest completed song.

Voices

Use list_voices when you need Claude to find a voice by style, tone, category, or description before generating speech or dialog.

GET/v1/speech/voices

Auth: API keyBilling: Free readBehavior: Voice catalog searchScopes: analytics:read

Credits and usage

Use get_credits before spendful work and get_usage to review recent activity. The Wubble extension keeps usage summaries compact so Claude can reason over them reliably.

GET/v1/credits

Auth: API keyBilling: Free readBehavior: Allowance readScopes: analytics:read

GET/v1/usage

Auth: API keyBilling: Free readBehavior: Recent usage summaryScopes: analytics:read

Included and deferred tools

Source-upload and local-file tools are not included in the first submitted bundle. The deferred tools are:

speech_to_textisolate_voicechange_voiceupload_audioupload_vocaldub_audioget_dubbing_transcriptanalyze_lyricscreate_podcast

Review-safe first bundle

speech_to_text, isolate_voice, change_voice,upload_audio, and related upload workflows are not included in the first submitted bundle. They can be revisited after a separate source-audio design.

Was this page helpful?

TroubleshootingNext