More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Drama Studio MCP — Local Speech & Directed Retakes

    1

    Generate speech on your Apple silicon Mac, choose voices and direct retakes using Drama Studio's separately installed local MCP server.

    Free

    Security scanned

    About this skill

    A free companion skill for agents connected to Drama Studio. It guides status checks, voice selection, local WAV generation, directed retakes and consent-based voice library management.

    This listing supplies instructions only: install Drama Studio and connect its local stdio drama-mcp server separately. Loading this skill through Agensi does not install or connect the server. Speech generation requires an Apple silicon Mac, macOS 26 or later, an installed Apple on-device speech model for the script language, the Drama 3 model weights, codec and recognition assets. Allow about 49 GB free disk plus generated audio; 64 GB or more unified memory is recommended, not a verified minimum. Model weights are governed by the separate Fish Audio Community License.

    The MCP server calls the app's loopback HTTP API and reads its local bearer token automatically. There is no hosted speech service or cloud speech API key included. Audio stays on the configured Mac unless you separately choose to share it. The six MCP tools do not provide live streaming, document import or project editing.

    Source and setup: https://github.com/maralski/drama-studio#agents-and-programs

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.

    Instructions only. Requires separately installed Drama Studio and a local stdio MCP connection on Apple silicon macOS 26+, with model weights, codec and recognition assets. About 49 GB free disk; 64 GB+ unified memory recommended, not a verified minimum. User must accept the separate Fish Audio Community License. No MCP live streaming, document import, project editing or multi-speaker casting tools.

    How to use

    1. 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. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 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

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 12 days ago

    • Free with an Agensi account

    Mac only

    Software
    Drama StudioApple siliconDrama MCPFish Audio Drama 3Apple on-device speech model

    drama-mcp runs locally over stdio. Fish Audio Drama 3 needs its weights and codec. Speech recognition uses the bundled assets.

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files

    Uses the separately configured local stdio MCP server. That server calls Drama Studio on loopback HTTP and reads its local bearer token automatically. It reads chosen reference clips and writes WAV audio; playback is optional. The companion skill requires no cloud speech account. Never upload the local API token.

    Listed12 days ago

    What's inside

    Frequently Asked Questions

    The connected Drama Studio MCP server generates WAV audio locally on your Mac through the app HTTP API. This skill supplies workflow instructions for that server; it does not include a hosted speech API.

    Speech generation requires an Apple silicon Mac, macOS 26 or later, an installed Apple on-device speech model for the script language, Drama Studio, its local stdio drama-mcp server, Drama 3 weights, codec and recognition assets. Allow about 49 GB free disk plus audio output. 64 GB or more unified memory is recommended, not a verified minimum.

    A SKILL.md instruction file covering status checks, voice selection, synthesis, directed retakes and consent-based voice library management. The app, executable, model assets and local MCP configuration are separate requirements.

    Yes. The separately installed MCP server can list the local voice library and add a voice from a local audio clip and transcript. Clone a real speaker only with their consent.

    These six MCP tools support file-based speech and retakes. They do not provide live streaming, document import, project editing or multi-speaker casting. Word verification does not guarantee acting quality or perfect performance cues.

    New in Agents & Orchestration

    Free