Skip to main content
Glama
Sign In
cursor-tts-mcp

cursor-tts-mcp

by LrTox
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

cursor-tts MCP

Progress-triggered TTS for Cursor Agent (speak / stop). Voice follows the system UI language (edge-tts; Windows-first).

Marketplace / Plugin install

Requires uv (uvx).

Way

Steps

Cursor Marketplace

After listing: search cursor-tts in Cursor

Local Plugin

Clone repo → Cursor: add local plugin / copy into ~/.cursor/plugins/local/cursor-tts

Manual MCP

Point MCP at uvx --from <this-repo> cursor-tts-mcp (see root mcp.json)

Plugin layout:

  • .cursor-plugin/plugin.json

  • mcp.json (uvx --from ${PLUGIN_ROOT})

  • rules/cursor-tts-speak.mdc

  • assets/logo.svg

Submit checklist: docs/guide/Marketplace提交清单.md · Form: https://cursor.com/marketplace/publish

Related MCP server: mcp-tts

Docs

See docs/README.md.

Quick start (dev)

cd E:\WorkSpace\mcp
python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -e ".[dev]"
pytest

Default Marketplace mode is embedded (MCP process speaks). Optional P1: tts_service on 127.0.0.1:18765 + thin HTTP client.

# optional standalone service
python -m tts_service

Local Cursor wiring:

  • .cursor/mcp.json → MCP cursor-tts

  • .cursor/rules/cursor-tts-speak.mdc → speak-on-progress rule

If tools missing: reload MCP / restart Cursor. See docs/guide/Cursor接入与配置说明.md.

Layout

.cursor-plugin/       # Marketplace plugin manifest
mcp.json              # Portable MCP launch for Plugin
rules/                # Agent rule shipped with Plugin
src/cursor_tts_mcp/   # MCP + orchestrator + engines + playback
config/default.yaml   # defaults
docs/                 # design & contracts
tests/

Available Tools

2 tools
speakA

Speak a short status line for the user. Use for real progress, task conclusions, errors/blocks, and when the user must decide or act. Always call this when delivering a conclusion — do not only type. Do not read code or long text. Do not wait until the whole task finishes for the first call. Speak in the system UI language (en): use short English phrases (≤80 characters). Do not force Chinese if the system language is not Chinese.

Inputs: text (required), category, priority, interrupt. Returns: result

stopA

Immediately stop current TTS playback. Default clear_queue=true; if false, only stop current and keep the queue.

Inputs: clear_queue. Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

    TDQS

    A4.5/5.0

    Scored across 2 tools

    Disambiguation5/5

    The two tools have completely distinct purposes: speak outputs audio, stop halts playback. No overlap or ambiguity exists between them.

    Naming Consistency5/5

    Both tool names are single-word imperative verbs (speak, stop) that directly describe their actions. The naming pattern is perfectly consistent and predictable.

    Tool Count4/5

    With only two tools, the set is minimal but appropriate for a narrow TTS status-line server. It feels slightly thin, but each tool earns its place.

    Completeness5/5

    For the stated purpose of speaking short status lines and stopping playback, the tools cover the full lifecycle with no obvious gaps. Stop even offers queue-clearing behavior.

    Maintenance

    ActivityStale
    ResponsivenessNo recent community issues

    Resources

    Related MCP Connectors

    Related MCP Servers

    • Yak MCP

      thamaji
      F
      license
      Not graded
      quality
      B
      maintenance
      Enables coding agents to speak aloud using text-to-speech functionality. Works with agents running inside devcontainers and provides configurable voice settings for creating chatty AI companions.
      6
      -

    • mcp-tts

      EnviralDesign
      A
      license
      Not graded
      quality
      D
      maintenance
      Adds text-to-speech capabilities to Cursor IDE, allowing your AI assistant to speak responses, summaries, and explanations out loud using OpenAI or ElevenLabs.
      1
      Apache 2.0

    • agent-voice

      doer1296
      A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI coding assistants to verbally announce task progress and events through local text-to-speech, with cloud synthesis via Volcano Engine or Edge TTS, scene-based voice settings, and a Windows SAPI fallback so announcements continue even when the cloud fails.
      MIT