Sag

ElevenLabs text-to-speech with mac-style say UX.

Security Scan

The skill's runtime instructions match an ElevenLabs TTS CLI, but the registry metadata (no envs/install) conflicts with SKILL.md (which requires ELEVENLABS_API_KEY and suggests installing a third‑party brew formula), so the package metadata is inconsistent and needs verification before installation.

Purpose & Capability

The SKILL.md clearly implements an ElevenLabs TTS wrapper (requires a sag CLI and an ELEVENLABS_API_KEY), which is coherent with the skill name/description. However the registry-level requirements claimed 'none' for env vars and install, which conflicts with SKILL.md metadata that lists bins and ELEVENLABS_API_KEY — this mismatch is suspicious (likely packaging/metadata error) and should be resolved.

Instruction Scope

The instructions are narrowly scoped to generating TTS via the sag CLI, writing an output file (e.g., /tmp/voice-reply.mp3), and returning it. They do not instruct reading unrelated files or exfiltrating other environment variables. They reference additional optional env names (SAG_API_KEY, ELEVENLABS_VOICE_ID/SAG_VOICE_ID) which are relevant to TTS usage.

Install Mechanism

The SKILL.md metadata suggests installing via a Homebrew formula from the steipete/tap (third‑party) tap. A Homebrew formula is an expected install path for a CLI, but using a third‑party tap carries some risk — and the registry metadata omitted this install spec, creating an inconsistency that should be checked against the homepage/source.

Credentials

Requested credentials (ELEVENLABS_API_KEY; optional SAG_API_KEY and voice ID vars) are proportional to an ElevenLabs TTS client. There are no unrelated credential requests. The inconsistency is that registry metadata declares no required env vars while SKILL.md expects an API key.

Persistence & Privilege

The skill does not request always:true or any elevated/persistent platform privileges. It's user-invocable and uses the CLI at invocation time; autonomous invocation remains possible (platform default) but is not combined with other high-risk flags.

What to consider before installing

Before installing or providing an API key: (1) Verify the source code/homepage (https://sag.sh) and that the Homebrew formula (steipete/tap/sag) is legitimate; (2) Favor creating a scoped ElevenLabs API key (minimal permissions) if possible; (3) Be cautious about installing a CLI from a third‑party Homebrew tap — review the formula and its upstream repo; (4) Note the registry metadata and SKILL.md disagree about required env/install information — ask the publisher to correct metadata or provide the source; (5) If you do install, confirm the sag binary's behavior locally before supplying secrets to an agent, and only grant the ELEVENLABS_API_KEY to environments you trust.

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

# 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)