A command for starting an ongoing voice conversation with the user, optionally choosing a voice and an opening message.

Use it to start a voice session, choose a supported voice, provide an initial spoken message, and manage turn-taking when multiple agents share the conversation.

It provides a voice-based way to continue a conversation instead of exchanging only written messages.

converse — Skill for Claude Code and Codex by mbailey: A command for starting an ongoing voice conversation with…

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add skills/mbailey/voicemode/converse

Any agent

npx skills add mbailey/voicemode --skill converse

Clone the repo

git clone --depth 1 https://github.com/mbailey/voicemode

Made for: Claude Code, Codex.

Or install voicemode, the plugin that ships this one along with the rest of its 5 skills, 3 commands, 1 agent, 5 hooks, 1 MCP server.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/mbailey/voicemode/converse.svg)](https://agentmods.dev/skills/mbailey/voicemode/converse)

Your own site

<a href="https://agentmods.dev/skills/mbailey/voicemode/converse"><img src="https://agentmods.dev/badge/skills/mbailey/voicemode/converse.svg" alt="Measured on agentmods" height="20"></a>

Per session 8 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 747 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00008	$0.00747
Opus 5	$0.00004	$0.00374
Sonnet 5	$0.00002	$0.00149
Haiku 4.5	$0.00001	$0.00075

Measured 4d ago against content hash 12b84f1a0198, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

converse scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/converse/SKILL.md · 75 lines

How it starts

The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/voicemode:converse

Start an ongoing voice conversation with the user using the voicemode:converse MCP tool.

Implementation

Call the voicemode:converse MCP tool. Argument handling:

$1 — optional voice name (e.g. af_river, samantha, nova). If non-empty, pass it as the voice parameter to the tool. If empty, omit voice so the tool uses its default.
$2 — optional initial message. If non-empty, use it as the message to speak. If empty, let the tool / Claude choose an appropriate opener.
To discover voices, read the voice://voices MCP resource; to speak in character, first read the persona at ~/.voicemode/voices/<voice>/README.md (index: ~/.voicemode/voices/PERSONAS.md).

All other parameters have sensible defaults.

Multi-agent turn-taking: in a session where more than one agent shares the voice channel, pass hold_conch=true when your next converse call will continue the thread (you're asking a question you'll answer, or speaking across several turns). It keeps the floor so other agents wait instead of cutting in at the turn boundary; the default false releases it after the exchange.

Examples

/voicemode:converse — no args, default voice, Claude chooses opener
/voicemode:converse af_river — use voice af_river, Claude chooses opener
/voicemode:converse af_river "let's plan the day" — use voice af_river, open with the given message

If $1 is provided but doesn't look like a known voice name (e.g. it contains spaces or punctuation that voices don't have), assume the user meant it as a message and pass it as the message instead — don't fail silently on a typo.

If MCP Connection Fails

If the MCP server isn't connected or the tool isn't available:

Run the install command:
```
/voicemode:install
```
This installs VoiceMode CLI, FFmpeg, and local voice services.

Or install manually via CLI:

uvx voice-mode-install --yes
voicemode whisper service install
voicemode kokoro install

Read the full file on GitHub · 75 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

4d ago First seen · 75 lines · 8 tokens per session scan A 12b84f1a0198

Subscribe to this mod's changes

converse is a skill published in the GitHub repository mbailey/voicemode (1,347 stars, last pushed 3d ago), licensed MIT. It adds 8 tokens to every session and 747 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

anthropic asr claude claudecode kokoro livekit

baby-sit

Monitor a GitHub pull request until CI is green, diagnose failures, and rerun only evidence-backed flaky GitHub Actions jobs.

langchain-ai/open-swe · 30 tokens

continual-learning

Nightly refinement of an existing per-repo review-style prompt using this reviewer's own finding outcomes. Read confirmed (resolved-by-commit / thumbs-up) and dismissed (thumbs-down) findings, promote the bug patterns the team actually fixes, demote the false-positive patterns, reconcile against the current prompt…

langchain-ai/open-swe · 89 tokens

converse