elevenlabs/skills
9 skills25.6k installs2.9k starsGitHub
Install
npx skills add https://github.com/elevenlabs/skillsSkills in this repo
1Text To SpeechThe text-to-speech skill documents how solo builders integrate ElevenLabs voice AI to convert text into downloadable or streamable audio. It covers Python and JavaScript clients with streaming write-to-file examples, cURL for quick tests, voice and model selection including multilingual v2, and setup via an installation reference. Use it when you are adding narration, in-app voice responses, accessibility audio, or agent tools that speak results in seventy-plus languages. You need an ElevenLabs API key in the environment and outbound network access; the skill steers JavaScript users toward official @elevenlabs packages only. It is a focused integration skill rather than a full voice-product architecture guide.5.8kinstalls2Speech To TextSpeech-to-text is an agent skill for solo builders who need production-ready ElevenLabs transcription in JavaScript, TypeScript, Python, or raw HTTP. It walks through installing the official `@elevenlabs/elevenlabs-js` client, configuring API keys safely via environment variables, and optional browser or React Scribe clients for live or file-based audio. Use it when you are adding voice input, meeting notes, or content repurposing to a SaaS feature, CLI tool, or agent workflow and want consistent package names instead of copy-paste from outdated tutorials. The readme emphasizes deprecating the old `elevenlabs` npm package and aligning imports with `ElevenLabsClient`, `Scribe`, and `useScribe` so your ship checklist does not break on the next dependency audit.4.2kinstalls3AgentsElevenLabs Agents is an agent skill for solo builders who want production-grade voice assistants without building speech stacks from scratch. It centers the ElevenLabs Agents Platform: install the CLI, authenticate, scaffold a project, pick a template such as customer-service or voice-only, then push configuration to ElevenLabs cloud. Python snippets show how to create agents programmatically with first messages, language, prompts, and versioning enabled—useful when your repo owns infrastructure as code. The skill fits when you are implementing interactive voice characters, support bots, or embedded site assistants where latency and natural turn-taking matter. You should already have an API key and internet access; the skill steers agent shape, conversation config, and deployment rhythm rather than generic LLM chat UI. After integration, you typically wire your product’s auth, analytics, and guardrails around the hosted voice session.3.9kinstalls4Sound EffectsSound-effects is an ElevenLabs integration skill for solo builders adding procedural or prompt-driven audio to apps, games, demos, or marketing clips without hiring a sound designer for every asset. It documents installation across JavaScript/TypeScript (`@elevenlabs/elevenlabs-js`), Python (`elevenlabs`), and raw REST/cURL, with a clear warning to avoid the legacy npm package. Authentication follows the standard ElevenLabs pattern: environment variable preferred, direct apiKey acceptable for local dev. The worked cURL example shows text-to-SFX generation written to `output.mp3`, which maps cleanly to agent workflows that scaffold a small script or fetch step during a build task. Complexity is beginner-friendly for anyone who already manages API keys. Use it when you need integration glue, not when you need mixing, licensing review, or offline asset libraries only.3.2kinstalls5MusicThe ElevenLabs music skill is an integration reference for generating music through the ElevenLabs Music API inside agent-assisted development. Solo builders adding soundtracks to demos, mobile experiences, or marketing clips can follow documented parameters for prompt-based composition, detailed plans, uploads, and video-to-music conversion without re-reading scattered SDK docs. The readme centers on `compose` streaming audio, optional `music_v1` model selection, instrumental forcing, and section duration enforcement when using composition plans. It suits indie developers who already have API keys and need copy-paste client patterns in Python or JavaScript plus guardrails on required fields and mutual exclusivity between prompt and plan modes. Use during feature implementation when your product needs programmatic music, not when you only need manual DAW production.3.2kinstalls6Setup Api KeySetup-api-key is a guided workflow for solo builders wiring ElevenLabs into Claude-style MCP stacks. It always checks for an existing ELEVENLABS_API_KEY in the environment or .env, validates against the ElevenLabs user endpoint, and only runs the full onboarding flow when the key is missing or invalid—reducing repeated prompts when voice tooling is already live. When setup is required, it points to the official API keys page and helps place the secret where MCP servers expect it. Use when ElevenLabs tools fail for auth, when onboarding a new machine, or when rotating credentials. Requires internet access to elevenlabs.io and api.elevenlabs.io per skill metadata. It does not generate audio or manage billing; it unblocks the credential layer so downstream ElevenLabs skills and MCP tools can run.3.1kinstalls7Voice Isolatorvoice-isolator is an agent skill that walks solo builders through integrating ElevenLabs Voice Isolator—background noise removal and vocal/speech isolation—using the supported JavaScript and Python SDKs or the REST API. Install it when you are building podcasts tools, voice agents, dubbing pipelines, or mobile/desktop apps that need cleaner input before transcription, TTS, or publishing. The skill stresses the current `@elevenlabs/elevenlabs-js` package, environment-variable-based keys, and a concrete cURL isolation request so agents do not resurrect deprecated clients. It fits the Build integrations moment: you already have noisy recordings and need a repeatable API call pattern, error handling, and secure key storage. Pair with ElevenLabs setup skills for onboarding keys before batch-processing assets in content or agent products.890installs8Voice ChangerVoice Changer is an ElevenLabs integration skill for solo builders who need speech-to-speech conversion inside agents, SaaS demos, or content pipelines. It walks through installing the current `@elevenlabs/elevenlabs-js` or Python `elevenlabs` client, authenticating with an API key, and calling the speech-to-speech API with a source audio file and target voice id. The readme emphasizes not using the legacy npm package, using environment variables for secrets, and matching the documented cURL shape for quick verification. Use it when you are past prototyping and need a repeatable, copy-paste integration path so Claude Code or Cursor can generate correct client code without guessing endpoints or headers. It pairs naturally with other ElevenLabs skills for keys and voices but stays focused on the STS conversion path and output formats such as mp3.744installs9Speech EngineSpeech Engine is an ElevenLabs integration skill for solo builders adding conversational voice to web, Next.js, Electron, or agent-backed apps. It walks through server environment variables, dependency installs, and a browser token pattern so API credentials never ship to client bundles. The readme emphasizes ngrok for local WebSocket exposure during development and calls out a temporary LiveKit WebRTC compatibility workaround via package overrides when sessions fail on /rtc/v1 paths. Use it when you have chosen ElevenLabs Speech Engine and need a disciplined install checklist rather than ad-hoc snippets. It does not replace product design for turn-taking, privacy disclosures, or production scaling—you still own auth, rate limits, and monitoring after the integration compiles.573installs