How do I install the Tts Tool skill?

Run npx skillmds@latest add planetis-m/tts-tool in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Tts Tool safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Tts Tool?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Tts Tool free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Tts Tool?

planetis-m (@planetis-m) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Tts Tool by planetis-m — Agent Skill

Name: Tts Tool
Rating: 4.666666666666667 (3 reviews)
Author: planetis-m

Tts Tool

Generate a natural-sounding `.opus` audio file from text or markdown. Use when a user wants spoken audio, especially for content with markdown, URLs, formulas, paths, or technical notation that should be rewritten for speech.

planetis-m Updated Sep 22, 2026

File contents

TTS Tool

Generate one .opus file with:

chunktts INPUT.txt OUTPUT.opus

Readiness

Run chunktts --help.
If chunktts is missing or fails to start, read references/chunktts-install.md, install the binary or missing runtime dependency, and retry.
After installing chunktts, tell the user how to configure DEEPINFRA_API_KEY or config.json before generating audio.

Execution

Request unrestricted network or escalated execution directly when running chunktts. Do not run a sandboxed network probe first.
Do not inspect environment variables, shell profiles, or filesystem files to discover API keys. If authentication fails, ask the user to configure the key, then retry.

Rewrite for Speech

Read the source directly and manually rewrite it for speech. Do not use a programmatic cleanup pipeline. Preserve the content while optimizing for listening rather than visual fidelity.

Remove markdown syntax while keeping useful visible text.
Turn bullet points and numbered items into normal spoken sentences.
For markdown links, keep the visible text and drop the URL unless it matters.
Rewrite raw URLs, email addresses, paths, flags, and technical identifiers into natural spoken forms. Avoid punctuation-heavy text.
Rewrite email addresses into forms like name at domain dot com.
Remove LaTeX delimiters and speak common math notation, such as \alpha as alpha, = as equals, / as over, and ^2 as squared.
Drop decorative or purely visual content that adds no spoken value.

Chunking

Insert <bk> between spoken thought units. Use paragraphs and list items as the default units; keep adjacent short sentences together when they form one thought.
Keep headings only when they add spoken meaning.
Do not summarize or drop meaningful content to fit a chunk. Rewrite dense sentences into shorter spoken sentences first.
If a paragraph or list item is too long, split it first at sentence boundaries.
Split inside a sentence only as a last resort.
Target 260-420 characters per chunk. Prefer at most 520; never exceed 620.
Remove empty chunks and collapse repeated <bk> markers.

Produce Output

Write the cleaned text with <bk> markers to a temporary input file.
Run chunktts against that file and an agent-managed .opus output path.
Tell the user the exact location of the generated .opus file.

---name: tts-tooldescription: Generate a natural-sounding `.opus` audio file from text or markdown. Use when a user wants spoken audio, especially for content with markdown, URLs, formulas, paths, or technical notation that should be rewritten for speech.---# TTS ToolGenerate one `.opus` file with:```bashchunktts INPUT.txt OUTPUT.opus```## Readiness- Run `chunktts --help`.- If `chunktts` is missing or fails to start, read [references/chunktts-install.md](references/chunktts-install.md), install the binary or missing runtime dependency, and retry.- After installing `chunktts`, tell the user how to configure `DEEPINFRA_API_KEY` or `config.json` before generating audio.## Execution- Request unrestricted network or escalated execution directly when running `chunktts`. Do not run a sandboxed network probe first.- Do not inspect environment variables, shell profiles, or filesystem files to discover API keys. If authentication fails, ask the user to configure the key, then retry.## Rewrite for SpeechRead the source directly and manually rewrite it for speech. Do not use aprogrammatic cleanup pipeline. Preserve the content while optimizing forlistening rather than visual fidelity.- Remove markdown syntax while keeping useful visible text.- Turn bullet points and numbered items into normal spoken sentences.- For markdown links, keep the visible text and drop the URL unless it matters.- Rewrite raw URLs, email addresses, paths, flags, and technical identifiers into natural spoken forms. Avoid punctuation-heavy text.- Rewrite email addresses into forms like `name at domain dot com`.- Remove LaTeX delimiters and speak common math notation, such as `\alpha` as `alpha`, `=` as `equals`, `/` as `over`, and `^2` as `squared`.- Drop decorative or purely visual content that adds no spoken value.## Chunking- Insert `<bk>` between spoken thought units. Use paragraphs and list items as the default units; keep adjacent short sentences together when they form one thought.- Keep headings only when they add spoken meaning.- Do not summarize or drop meaningful content to fit a chunk. Rewrite dense sentences into shorter spoken sentences first.- If a paragraph or list item is too long, split it first at sentence boundaries.- Split inside a sentence only as a last resort.- Target `260-420` characters per chunk. Prefer at most `520`; never exceed `620`.- Remove empty chunks and collapse repeated `<bk>` markers.## Produce Output- Write the cleaned text with `<bk>` markers to a temporary input file.- Run `chunktts` against that file and an agent-managed `.opus` output path.- Tell the user the exact location of the generated `.opus` file.

Frequently asked questions