> ## Documentation Index
> Fetch the complete documentation index at: https://docs.slng.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent skill

> Install the slng-ai/skills pack so a coding agent can call the SLNG API: speech to text, text to speech, and voice agents.

`slng-ai/skills` is a pack of skills for coding agents. Once installed, your
coding agent knows how to set up an SLNG key, synthesize speech, transcribe
audio, draft a voice-agent prompt, and dispatch outbound calls. Each skill is a
short markdown brief plus example code the agent runs directly, so you describe
the task in plain language instead of writing the client.

For the migration skills that move an existing LiveKit, Pipecat, or custom agent
onto SLNG, see the [Integrate skill](/integrations/skills/integrate-skill).

## Install

The `skills` CLI needs Node 18 or later. It detects the coding agents you have
installed, or prompts you to pick one.

```bash theme={null}
npx skills add slng-ai/skills
```

Skills install into your coding agent's skills directory, for example
`~/.claude/skills/`. Re-run the command any time to update.

* Target one coding agent with `-a` (`claude-code`, `codex`, or `gemini-cli`).
* Add `-g` to install for every project.
* Name a single skill to install just that one:

```bash theme={null}
npx skills add slng-ai/skills text-to-speech
```

## Set up your key

The core skills read your key from `VOICEAI_API_KEY`:

```bash theme={null}
export VOICEAI_API_KEY="your-slng-api-key"
```

Create a key at [app.slng.ai](https://app.slng.ai/api-keys), or see
[Create your API key](/guides/get-started/quickstart#create-your-api-key). The
`setup-api-key` skill configures and validates this on first use.

## Skills in the pack

| Skill            | What the agent learns to do                                          |
| ---------------- | -------------------------------------------------------------------- |
| `setup-api-key`  | Configure and validate `VOICEAI_API_KEY`.                            |
| `text-to-speech` | Generate speech from text with any supported model and voice.        |
| `speech-to-text` | Transcribe audio to text.                                            |
| `agent-prompt`   | Draft a voice-agent prompt for the SLNG Agent Builder.               |
| `agents`         | Create, manage, and dispatch voice agents, including outbound calls. |

Each skill shows the same task four ways: the `voiceai` CLI, the Python SDK, the
JavaScript SDK, and raw REST or WebSocket calls. The agent picks whichever fits
the project it is working in. For the full model list, see the catalog for
[text to speech](/models/catalog/text-to-speech) and
[speech to text](/models/catalog/speech-to-text).

## Use a skill

Once a skill is installed, describe the task to your coding agent. The skill's
description tells the agent when to load it, so you do not call it by name.

```text theme={null}
Transcribe recording.wav with SLNG and summarize the result.
```

The agent loads the `speech-to-text` skill, runs `voiceai stt recording.wav` (or
the SDK equivalent when the project is already Python or JavaScript), parses the
response, and writes the summary back to you.

## Works with

Any coding agent that loads skills from a directory: Claude Code, Codex, Gemini
CLI, and other compatible hosts. The skills are plain markdown plus
shell-runnable examples.

## Next steps

<CardGroup cols={2}>
  <Card title="Skills guide" icon="book-open" href="/guides/integrate/skills">
    A walkthrough of the pack, with example prompts for each skill.
  </Card>

  <Card title="Integrate skill" icon="arrow-right-left" href="/integrations/skills/integrate-skill">
    Migration skills that move an existing agent onto SLNG.
  </Card>

  <Card title="GitHub repo" icon="github" href="https://github.com/slng-ai/skills">
    Source, version history, and a place to propose new skills.
  </Card>
</CardGroup>