Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+21 moreClaude CodeCursorCodex CLIMuseDots
Manus
Grok
Grok Bot+18 agentsSpoken Audio Forge
Turn scripts, CSVs, and articles into professional, reproducible spoken audio — narration, voiceovers, and IVR prompts with steerable tone and pacing.
Free
Spoken Audio Forge
Example session with this skill installed
I have a product walkthrough script in docs/onboarding-walkthrough.md and 14 short support prompts in support-lines.csv. Please turn the walkthrough into numbered MP3 narration segments with a calm, trustworthy tone, and render each support line as its own WAV file with a clear, unhurried phone-menu delivery. Use the cedar voice for the walkthrough and marin for the support prompts, keep every chunk under the per-request character limit, and drop the final files into the standard output folder with a record of the exact settings used for each clip.
- Read your context and instructions
- Compiled the spoken audio forge
Summary
Rendered the onboarding walkthrough as 4 continuous MP3 segments and all 14 support lines as WAV prompts; manifests cleaned and per-clip settings logged.
| Item | Requested | Delivered |
|---|---|---|
| Walkthrough segments | 1 script, calm tone | walkthrough-001.mp3 … walkthrough-004.mp3 (cedar, moderate pacing) |
| Support prompts | 14 CSV lines | prompt-001.wav … prompt-014.wav (marin, measured pacing, pauses between options) |
| Rate limit | batch-safe | 18 requests at 45 RPM, 0 failures |
| Longest chunk | under per-request cap | 3,812 chars, split at a section boundary |
Next steps
- Listen to walkthrough-002 — the acronym "SPRV" may need a pronunciation note
- If the support prompts feel rushed, change only pacing and re-render the batch
- Files are in output/spoken-audio/; keep the instruction specs for future re-renders
- Disclose synthetic voice wherever this audio is presented as a human narrator
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Generating high-quality, professional text-to-speech for product demos, IVR systems, or narration often requires tedious manual tweaking and custom scripts. Developers face challenges with inconsistent pacing, fragmented batch processing, and hit-or-miss tone control across multiple files.
What it does
- Synthesizes scripts, CSVs, or articles into professional audio clips using OpenAI Audio API models.
- Handles long-form content by automatically segmenting text into chunks under 4096 characters to prevent truncation.
- Executes batch jobs via JSONL manifests for consistent generation of multiple prompts or lines.
- Applies steerable delivery specs to control affect, tone, pacing, and diction for specific contexts like accessibility or sales.
- Manages local file organization with stable, sortable naming conventions for reproducible workflows.
Frameworks & tools
OpenAI Audio API (GPT-4o mini TTS), Python openai library, and uv for package management.
Why this beats prompting it yourself
This skill provides a structured CLI-first approach that ensures auditability and reproducibility. It avoids the common pitfalls of ad-hoc scripts by enforcing hard constraints on character limits and API rate-limiting while providing standardized templates for tone and pacing that go beyond basic text prompts.
Use cases
- Creating voiceovers for product walkthroughs and explainer videos.
- Generating numbered IVR prompt sets for phone menus and support systems.
- Producing accessibility readouts for long-form articles or documentation.
- Batch processing localized UI strings into spoken notification clips.
Known limitations
Does not support voice cloning, custom voice enrollment, or real-time streaming. Limited to built-in OpenAI voices. No post-production editing or music mixing capabilities.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.
Synthesis only — no voice cloning, custom enrollment, streaming conversation, music, sound effects, or transcription. Live calls require your own OpenAI key, bill per request, and are limited to built-in voices.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
4 uses
Times used to date
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account
- Accounts
OpenAI
- Software
Python
Permissions
Allowed Hosts
File Scopes
Needs shell and project file access to run the bundled CLI and organize tmp/output audio files, plus network and OPENAI_API_KEY to call the OpenAI Audio API for live synthesis.