-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Voice notifications for AI coding agents, powered by hook-driven TTS.
Jarvis Line speaks short progress, completion, permission, and input-request summaries so you can follow long-running agent work without watching every session. A single local audio worker keeps simultaneous sessions from talking over each other.
Jarvis line: The requested change is complete.
Current release:
v0.8.3. Jarvis Line is beta software. The macOS CLI/runtime is the primary validated surface. The macOS app, Windows, and Linux are Preview. Wiki pages may describe features already merged for the next tag; tagged installs only contain features listed in that release's notes.
python3 -m pip install "git+https://github.com/reitenji/jarvis-line.git@v0.8.3"
jarvis-line setup
jarvis-line doctor
jarvis-line tts test --text "Jarvis line test is ready."Setup asks you to choose a spoken language, compatible TTS backend, speech mode, agent, instruction scope, and optional integrations. Jarvis Line prints the instruction block for you to review and paste into the Markdown file your agent actually reads; it does not modify that file by default.
Continue with Getting Started.
| Event | Example |
|---|---|
| Commentary | "The test suite is running." |
| Final | "The implementation is complete." |
| Permission | "Permission is needed to push changes." |
| Input required | "Your release choice is required." |
Permission and input-required speech is optional. See Attention Alerts for availability, privacy, and queue behavior.
Final speech has a short notification chime by default in v0.8.3, making a
completed task distinguishable from commentary. The cue can be disabled or
given its own volume without changing speech volume.
| Guide | Use it for |
|---|---|
| Getting Started | Install, configure, verify, and update |
| Attention Alerts | Permission and input-request speech |
| Agent Instructions |
AGENTS.md, CLAUDE.md, and GEMINI.md
|
| TTS and Voices | Kokoro, system voices, languages, and custom TTS |
| Configuration | Safe settings, profiles, and config locations |
| CLI Reference | Command map and common examples |
| Troubleshooting | No speech, hook, queue, and diagnostic checks |
| macOS App | Native manager and DMG status |
| Platform and Agent Support | Validated and Preview surfaces |
| Privacy and Diagnostics | Local data, network activity, and issue reports |
- Ask usage questions in Q&A Discussions.
- Propose features in Ideas.
- Report reproducible defects with the bug report form.
- Report vulnerabilities privately through GitHub Security Advisories.
Repository · Releases · Issues · Discussions