speak

speak is a skill for Claude Code, Codex from fakoli/fakoli-plugins. It costs 33 tokens per session (476 once invoked), scanned A, original, MIT.

A text-to-speech workflow that sends requested text to a selected speech provider and plays the result. It can inspect available providers, voices, playback support, and estimated cost, and can manage optional automatic speech.

In plain words
What is it for?
Use it to read text aloud, inspect speech voices and providers, check estimated cost, or control automatic speech.
Why use it?
It avoids manually checking speech settings or building playback commands. It also makes provider, voice, cost, and privacy considerations visible before text is sent.

Skill for Claude CodeCodex

Written for Claude Code and Codex: shipped in a Claude Code plugin, but also agents/openai.yaml present. Also seen: mentions Codex.

Part of the fakoli-speak plugin — 1 skill, 7 commands shipped together

Good fit Use it to read text aloud, inspect speech voices and providers, check estimated cost, or control automatic speech.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/fakoli/fakoli-plugins/speak
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.

Any agent
npx skills add fakoli/fakoli-plugins --skill speak
Clone the repo
git clone --depth 1 https://github.com/fakoli/fakoli-plugins

Made for: Claude Code, Codex.

Or install fakoli-speak, the plugin that ships this one along with the rest of its 1 skill, 7 commands.

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.

agentmods badge for speak

README.md
[![agentmods](https://agentmods.dev/badge/skills/fakoli/fakoli-plugins/speak/github.svg)](https://agentmods.dev/skills/fakoli/fakoli-plugins/speak)
Your own site
<a href="https://agentmods.dev/skills/fakoli/fakoli-plugins/speak"><img src="https://agentmods.dev/badge/skills/fakoli/fakoli-plugins/speak/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for speak

Your own site · 80×15
<a href="https://agentmods.dev/skills/fakoli/fakoli-plugins/speak"><img src="https://agentmods.dev/badge/skills/fakoli/fakoli-plugins/speak.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 33 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 476 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00033 $0.00476
Opus 5 $0.00016 $0.00238
Sonnet 5 $0.00007 $0.00095
Haiku 4.5 $0.00003 $0.00048

Measured 4d ago against content hash df8022693bdd, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

speak 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.

plugins/fakoli-speak/skills/speak/SKILL.md · 19 lines

What it actually says

Fakoli Speak

Resolve the plugin root from this installed skill (../..) and invoke uv run --frozen --directory "$PLUGIN_ROOT" fakoli-speak COMMAND. The CLI loads ~/.env without overwriting existing environment variables; never print keys. It requires Python, uv, and a supported player (afplay, mpv, or ffplay).

Use status, provider, voices, and cost --json to inspect the relevant configuration. The provider NAME command validates and displays a provider; it does not persist a switch. Set FAKOLI_SPEAK_PROVIDER for the invocation or update the user's configuration only when requested. Read README for provider-specific keys and model/voice settings.

For requested speech, pass text through stdin or a structured argv list. Text sent to a cloud provider leaves the machine; use the selected provider and requested text. macOS say is local. The current facade caps speech at 4,000 characters (or the provider limit); report truncated output instead of claiming the entire text was spoken. If the user needs the full text, split it deliberately and sequence playback, respecting any cost/provider constraints.

stop targets only the verified owned playback worker. Do not use broad pkill commands or signal an unverified PID. Synthesis completing and playback being launched are different from audible playback being verified. Playback depends on the device/player and session's process lifetime.

Autospeak is opt-in through autospeak on/off. Turning it on permits subsequent eligible responses to be sent to the configured provider. The bundled Claude Stop hook uses last_assistant_message and ignores recursive/malformed events. Codex can use the skill/CLI directly; hook support and event fields vary by host, so do not promise automatic playback without checking runtime delivery. It does not read arbitrary transcript files.

Costs are local character-based estimates/overrides, not provider invoices; token/audio-priced models and free tiers may differ substantially. Verify provider pricing before a budgeting decision. Never treat a default zero estimate as proof of free usage.

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

  1. 4d ago First seen · 19 lines · 33 tokens per session scan A df8022693bdd

Subscribe to this mod's changes

speak is a skill published in the GitHub repository fakoli/fakoli-plugins (4 stars, last pushed 5d ago), licensed MIT. It adds 33 tokens to every session and 476 once invoked, about $0.0002 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 2026-09-06.

Related

Other skills, from other repositories

speech-generation

Generate text-to-speech audio using the ElevenLabs MCP server's texttospeech tool. Use when the user asks to create voiceovers, narration, audio from text, spoken dialogue, podcast-style audio, audiobook readings, announcements, or voice content. Supports voice selection by name or ID, multiple models (multilingual…

lukaskellerstein/claude-my-marketplace · 87 tokens

visual-planning

ALWAYS use this skill FIRST, before producing or fetching ANY media asset — an image, video/GIF, chart, graph, diagram, infographic, SVG/icon, music, or speech/voiceover — and before any stock-photo search. It is the mandatory pre-generation gate. For a SINGLE asset it runs a short brief — clarify the ask → confirm…

lukaskellerstein/claude-my-marketplace · 271 tokens

demo-video

Orchestrates building a narrated demo video of a project — reads the codebase, writes a storyboard, prepares deterministic app state, drives the UI with Playwright to record clips (web and Electron), generates ElevenLabs voiceover, reconciles measured durations into a timeline, and renders the final cut with Remotion.…

lukaskellerstein/claude-my-marketplace · 116 tokens

image-generation

Generate and work with images using the media-mcp server's generateimage tool. Use when the user asks to create images, generate visuals, make illustrations, create thumbnails, design graphics, produce concept art, create product mockups, or work with reference images. Supports multiple aspect ratios, resolutions up…

lukaskellerstein/claude-my-marketplace · 79 tokens

media-prompt-craft

Craft effective prompts for AI image/video generation and search queries for stock photo services. Translates design direction and styleguide choices into specific, style-consistent prompts. Use when the user asks to "find images for my site", "generate a hero image", "what kind of images should I use", "create image…

lukaskellerstein/claude-my-marketplace · 112 tokens

demo-app-prep

Gets a project into a demo-worthy, deterministic state before recording — boots the app, seeds realistic data, bypasses or scripts login, freezes clocks and random values, hides dev-only UI, and makes state resettable between takes. Use before capturing any demo video section, when a demo shows empty states or…

lukaskellerstein/claude-my-marketplace · 87 tokens