Adds voice output capability (OutLoud) to Claude Code. After you give a response, the user can press a speaker hotkey or type /speak (or use a terminal button binding) to hear the output spoken aloud using the local speaker plugin. The plugin uses a Stop hook to remember your last answer (and optional autoSpeak). Use this whenever the user mentions speaker, voice, read it, speak, hear output, TTS, OutLoud, or wants to listen while working.
Scanned 9/19/2026
Install to Claude Code
npx -y skills add rhishi99/OutLoud --skill speaker --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Speaker?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/rhishi99-speaker)More formats (shields.io, HTML) on the badges page.
---
name: speaker
description: >
Adds voice output capability (OutLoud) to Claude Code. After you give a response,
the user can press a speaker hotkey or type /speak (or use a terminal button binding)
to hear the output spoken aloud using the local speaker plugin.
The plugin uses a Stop hook to remember your last answer (and optional autoSpeak).
Use this whenever the user mentions speaker, voice, read it, speak, hear output, TTS, OutLoud, or wants to listen while working.
license: MIT
---
# Speaker (OutLoud)
You have a voice now.
## How it works for the user
- Every time you finish responding, a Stop hook quietly saves your last output.
- User sees a 🔊 icon (via statusLine) in the Claude Code terminal.
- To hear the output:
- Press the configured hotkey (recommended: Ctrl+Alt+S or Cmd+Shift+S)
- Type `/speak`
- Or manually run the speak script from another terminal / launcher
## For you (the agent)
- Keep your final answers speakable: short paragraphs + clear structure are great for listening.
- When user says "speak the last" or "read it out", remind them to use the hotkey or `/speak`.
- If the response contains large code, the speaker will say "[code block]" for those sections.
- You can also say at the end of an important reply: "You can press the speaker button / hotkey to hear this."
## Commands the user can use
- `/speak` — speak last captured response
- `/speak Some custom text here` — speak arbitrary text
## Configuration / Status
Users can show the speaker status in Claude Code's status bar by adding the status script to their settings (see README).
## Technical
The capture is automatic and silent. Nothing is sent over the network. Speech happens locally on the user's machine using native system voices (or configured TTS).
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!