Sag
Wire macOS-friendly ElevenLabs text-to-speech into agent or CLI workflows so spoken output feels like `say` with expressive voice control.
Overview
sag is an agent skill for the Build phase that integrates the sag CLI with ElevenLabs text-to-speech for local, say-like voice playback in agent workflows.
Install
npx skills add https://github.com/steipete/clawdis --skill sagWhat is this skill?
- ElevenLabs TTS via `sag` with mac-style quick commands (`sag "Hello there"`, `sag voices`)
- Model tiers: default expressive `eleven_v3`, stable `eleven_multilingual_v2`, fast `eleven_flash_v2_5`
- Pronunciation fixes via respelling, `--normalize auto`, and `--lang` bias for names and URLs
- v3 delivery tags (`[whispers]`, `[laughs]`, pause markers) plus v2/v2.5 SSML break support
- OpenClaw metadata: brew install steipete/tap/sag, requires `ELEVENLABS_API_KEY`
- Documents 3 model options: eleven_v3, eleven_multilingual_v2, eleven_flash_v2_5
- Brew install path: steipete/tap/sag with required ELEVENLABS_API_KEY
Adoption & trust: 1.9k installs on skills.sh; 378k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
You want your agent or script to speak naturally on macOS but do not want to hand-roll ElevenLabs API calls, voice IDs, and pronunciation edge cases.
Who is it for?
macOS-based agent builders using OpenClaw/Clawdis who need quick ElevenLabs TTS from the terminal.
Skip if: Windows/Linux environments without sag, offline-only audio needs, or teams that cannot use cloud TTS API keys.
When should I use this skill?
ElevenLabs text-to-speech with mac-style say UX via the sag CLI.
What do I get? / Deliverables
You configure `ELEVENLABS_API_KEY`, pick a model and voice, and run verified `sag` commands with normalization and delivery tags for reliable local audio.
- Working sag speak commands with chosen voice and model
- Documented normalization and delivery tag settings for repeatable phrasing
Recommended Skills
Journey fit
Voice output is integrated during Build when you connect external APIs and local binaries to your agent stack. sag is a third-party CLI integration (brew-installed) with API keys and playback—classic integrations work, not core app UI.
How it compares
A thin CLI integration skill—not a hosted voice widget SDK or a multi-speaker podcast editor.
Common Questions / FAQ
Who is sag for?
Solo builders on macOS wiring ElevenLabs voices into agents or shell automations via the sag binary.
When should I use sag?
During Build integrations when you add spoken feedback, alerts, or demo narration to an agent that already supports shell tools.
Is sag safe to install?
Review the Security Audits panel on this Prism page; sag needs network access and an API key—treat `ELEVENLABS_API_KEY` as a secret.
SKILL.md
READMESKILL.md - Sag
# sag Use `sag` for ElevenLabs TTS with local playback. API key (required) - `ELEVENLABS_API_KEY` (preferred) - `SAG_API_KEY` also supported by the CLI 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`)