Collections of skills for building with ElevenLabs
Start reading — 6 minnpx skills add elevenlabs/skillsMIT459 stars7,756 wordsUpdated 16 September 2026
Typeset in this issue’s own colours, resolved from the awesome-design-md registry.

Agent skills for ElevenLabs (opens in a new tab) developer products. These skills follow the Agent Skills specification (opens in a new tab) and can be used with any compatible AI coding assistant.
npx skills add elevenlabs/skills| Skill | Description |
|---|---|
| text-to-speech (opens in a new tab) | Convert text to lifelike speech using ElevenLabs’ AI voices |
| speech-to-text (opens in a new tab) | Transcribe audio files to text with timestamps |
| speech-engine (opens in a new tab) | Add real-time voice conversations to a custom LLM or chat agent |
| agents (opens in a new tab) | Build conversational voice AI agents |
| sound-effects (opens in a new tab) | Generate sound effects from text descriptions |
| music (opens in a new tab) | Generate music tracks using AI composition |
| voice-changer (opens in a new tab) | Transform the voice in an audio recording into a different target voice (speech-to-speech) |
| voice-isolator (opens in a new tab) | Remove background noise and isolate vocals/speech from audio |
| dubbing (opens in a new tab) | Dub audio/video into other languages while preserving the original speakers’ voices |
| setup-api-key (opens in a new tab) | Guide through obtaining and configuring an ElevenLabs API key |
All skills require an ElevenLabs API key. Set it as an environment variable:
export ELEVENLABS_API_KEY="your-api-key"Get your API key from the setup-api-key skill or use the ElevenLabs dashboard (opens in a new tab).
Most skills include examples for:
pip install elevenlabsnpm install @elevenlabs/elevenlabs-jsnpm install -g @elevenlabs/cli or brew install elevenlabs/tap/elevenlabs (wraps the REST API; reads ELEVENLABS_API_KEY automatically)JavaScript SDK Warning: Always use
@elevenlabs/elevenlabs-js. Do not usenpm install elevenlabs(that’s an outdated v1.x package).
See the installation guide in any skill’s references/ folder for complete setup instructions including migration from deprecated packages.
The evals/ directory contains trigger and functional evaluations for all skills.
# Run all evaluations (trigger + functional)
python3 evals/run_all.py -v
# Trigger evals only — tests if skills fire for the right queries (~3 min)
python3 evals/run_all.py --trigger-only -v
# Functional evals only — tests if skills produce correct output (~15 min)
python3 evals/run_all.py --functional-only -v
# Specific skills
python3 evals/run_all.py --skills text-to-speech agents -v
# Custom model (see `cursor-agent --list-models`)
python3 evals/run_all.py --model gpt-5.4-high -vResults are saved to evals/results/<timestamp>/ with a report.md summary and results.json for programmatic access.
Functional evals use an isolated cursor-agent workspace per test case (under that results tree); they do not modify skill sources under each skill’s directory.
Requires the Cursor Agent CLI (opens in a new tab) (cursor-agent on your PATH; override binary with CURSOR_AGENT) and Cursor authentication (cursor-agent login or CURSOR_API_KEY).
MIT
main, last pushed 16 September 2026.SKILL.md, not by matching a directory convention. 2 distinct layouts observed: .agents/skills/*/SKILL.md, */SKILL.md.h1 and no skipped levels:/elevenlabs/skills.md, and each skill at its own .md URL.