sag — Claude Code skill: ElevenLabs TTS like macOS say

Q: How do I install sag in Claude Code?

Run `npx skills add openclaw/openclaw --skill sag -a claude-code`. The binary itself is installed separately with `brew install steipete/tap/sag`.

Q: Which API key does sag require?

An ElevenLabs key. Set it as `ELEVENLABS_API_KEY` or `SAG_API_KEY`. A key file via `--api-key-file` is also supported.

What sag does

sag is a CLI wrapper over ElevenLabs Text-to-Speech. Its interface mirrors the macOS say command: sag "Hello there" generates audio and plays it locally.

The binary installs via Homebrew from the steipete/tap/sag tap. You need an ElevenLabs API key, set as ELEVENLABS_API_KEY or SAG_API_KEY. A key file works too via --api-key-file.

The skill knows three ElevenLabs models: eleven_v3 as the expressive default, eleven_multilingual_v2 for stability, and eleven_flash_v2_5 for speed. sag voices lists available voices and sag prompting shows model-specific tips.

Voice character comes from audio tags placed at the start of a line: [whispers], [shouts], [sings], [excited], [crying], and more. For v3, pauses use [pause], [short pause], [long pause] because SSML `` is not supported. When the user asks for a voice reply, the skill generates an mp3 and sends it to the chat.

Install it if you already use ElevenLabs and want the agent to speak. If voice output is not part of your workflow, this is just an extra dependency and an API key to manage.

What it helps with

Speak text from the terminal with one command
Generate a voice reply from the agent and send an mp3 to chat
Pick a voice and model for the task
Add emotion and pauses to speech with tags
Normalize pronunciation of numbers, units, and URLs

Example prompts for the agent

Speak: "Hello, this is a test message"
Generate a voice reply with the Clawd voice and send the mp3
Show the available voices

How to install sag

1Skills CLI (Vercel)

npx skills add openclaw/openclaw --skill sag -a claude-code

2Manually: put the files in place

~/.claude/skills/sag/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/sag/

1Skills CLI (Vercel)

npx skills add openclaw/openclaw --skill sag -a codex

2Manually: put the files in place

.agents/skills/sag/SKILL.md

For all projects use ~/.agents/skills/

1Skills CLI (Vercel)

npx skills add openclaw/openclaw --skill sag -a cursor

2Manually: put the files in place

.cursor/skills/sag/SKILL.md

Cursor also reads .agents/skills/ and .claude/skills/

1GitHub CLI

gh skill install openclaw/openclaw sag

2Manually: put the files in place

.github/skills/sag/SKILL.md

1Skills CLI (Vercel)

npx skills add openclaw/openclaw --skill sag -a gemini-cli

2Platform CLI

gemini skills install https://github.com/openclaw/openclaw --path skills/sag

Manually: put the files in place

.devin/skills/sag/SKILL.md

The legacy .windsurf/skills/ path still works

Skills CLI (Vercel)

npx skills add openclaw/openclaw --skill sag -a opencode

From the same set48

The whole set →

Questions about sag

How do I install sag in Claude Code?

Run npx skills add openclaw/openclaw --skill sag -a claude-code. The binary itself is installed separately with brew install steipete/tap/sag.

What does sag do?

It converts text to speech via ElevenLabs TTS with local playback. sag "text" generates and plays audio, while -o saves it as an mp3 file.

Which API key does sag require?

An ElevenLabs key. Set it as ELEVENLABS_API_KEY or SAG_API_KEY. A key file via --api-key-file is also supported.

Source text

SKILL.md · MIT

---
name: sag
description: "ElevenLabs text-to-speech with mac-style say UX."
homepage: https://sag.sh
metadata:
  {
    "openclaw":
      {
        "emoji": "🔊",
        "requires": { "bins": ["sag"] },
        "primaryEnv": "ELEVENLABS_API_KEY",
        "install":
          [
            {
              "id": "brew",
              "kind": "brew",
              "formula": "steipete/tap/sag",
              "bins": ["sag"],
              "label": "Install sag (brew)",
            },
          ],
      },
  }
---

# sag

Use `sag` for ElevenLabs TTS with local playback.

API key (required)

- `ELEVENLABS_API_KEY` (preferred)
- `SAG_API_KEY` also supported by the CLI
- Key file via `--api-key-file`, `ELEVENLABS_API_KEY_FILE`, or `SAG_API_KEY_FILE`

Quick start

- `sag "Hello there"`
- `sag speak -v "Roger" "Hello"`
- `sag voices`
- `sag prompting` (model-specific tips)

Model notes

- Default: `eleven_v3` (expressive)
- Stable: `eleven_multilingual_v2`
- Fast: `eleven_flash_v2_5`

Pronunciation + delivery rules

- First fix: respell (e.g. "key-note"), add hyphens, adjust casing.
- Numbers/units/URLs: `--normalize auto` (or `off` if it harms names).
- Language bias: `--lang en|de|fr|...` to guide normalization.
- v3: SSML `<break>` not supported; use `[pause]`, `[short pause]`, `[long pause]`.
- v2/v2.5: SSML `<break time="1.5s" />` supported; `<phoneme>` not exposed in `sag`.

v3 audio tags (put at the entrance of a line)

- `[whispers]`, `[shouts]`, `[sings]`
- `[laughs]`, `[starts laughing]`, `[sighs]`, `[exhales]`
- `[sarcastic]`, `[curious]`, `[excited]`, `[crying]`, `[mischievously]`
- Example: `sag "[whispers] keep this quiet. [short pause] ok?"`

Voice defaults

- `ELEVENLABS_VOICE_ID` or `SAG_VOICE_ID`

Confirm voice + speaker before long output.

## Chat voice responses

When the user asks for a "voice" reply (e.g., "crazy scientist voice", "explain in voice"), generate audio and send it:

```bash
# Generate audio file
sag -v Clawd -o /tmp/voice-reply.mp3 "Your message here"

# Then include in reply:
# MEDIA:/tmp/voice-reply.mp3
```

Voice character tips:

- Crazy scientist: Use `[excited]` tags, dramatic pauses `[short pause]`, vary intensity
- Calm: Use `[whispers]` or slower pacing
- Dramatic: Use `[sings]` or `[shouts]` sparingly

Default voice for Clawd: `lj2rcrvANS3gaWWnczSX` (or just `-v Clawd`)

sag — Claude Code skill: ElevenLabs TTS with macOS say UX

What sag does

What it helps with

Example prompts for the agent

How to install sag

From the same set48

Similar

Questions about sag

Source text

See also