ModelsAtlas
Explore
Atlas / MCP servers / kvadratni / Speech

SpeechCAUTION

mcp/kvadratni/speech
MCP server by kvadratni · Audited 2026-10-07
⚑ Report / re-scanSource ↗

Speech MCP: A Goose MCP extension for voice interaction with audio visualization

Verdict
CAUTION
Grade
B
Trust score
89 /100
Exposed tools
7 6r · 1w · 0d
Transport
—
License
MIT
Stars
83
01

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

A Goose MCP extension for voice interaction with modern audio visualization.

https://github.com/user-attachments/assets/f10f29d9-8444-43fb-a919-c80b9e0a12c8

Overview

Speech MCP provides a voice interface for Goose, allowing users to interact through speech rather than text. It includes:

  • Real-time audio processing for speech recognition
  • Local speech-to-text using faster-whisper (a faster implementation of OpenAI's Whisper model)
  • High-quality text-to-speech with multiple voice options
  • Modern PyQt-based UI with audio visualization
  • Simple command-line interface for voice interaction

Features

  • Modern UI: Sleek PyQt-based interface with audio visualization and dark theme
  • Voice Input: Capture and transcribe user speech using faster-whisper
  • Voice Output: Convert agent responses to speech with 54+ voice options
  • Multi-Speaker Narration: Generate audio files with multiple voices for stories and dialogues
  • Single-Voice Narration: Convert any text to speech with your preferred voice
  • Audio/Video Transcription: Transcribe speech from various media formats with optional timestamps and speaker detection
  • Voice Persistence: Remembers your preferred voice between sessions
  • Continuous Conversation: Automatically listen for user input after agent responses
  • Silence Detection: Automatically stops recording when the user stops speaking
  • Robust Error Handling: Graceful recovery from common failure modes with helpful voice suggestions

Installation

Important Note: After installation, the first time you use the speech interface, it may take several minutes to download the Kokoro voice models (approximately 523 KB per voice). During this initial setup period, the system will use a more robotic-sounding fallback voice. Once the Kokoro voices are downloaded, the high-quality voices will be used automatically.

⚠️ IMPORTANT PREREQUISITES ⚠️

Before installing

Read from source at commit dce2018e7b2bOBSERVED · 2026-10-07
02

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.

claude-code
claude mcp add speech-mcp -- uvx speech-mcp
claude-desktop
{
  "mcpServers": {
    "speech-mcp": {
      "command": "uvx",
      "args": [
        "speech-mcp"
      ]
    }
  }
}
03

Exposed tools (7)

6 read · 1 write · 0 destructive.

ToolRiskDescription
close_uiread
launch_uiread
narrateread
narrate_conversationread
replyread
start_conversationwrite
transcriberead
04

Trust audit

CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.

LayerWhat it checksResult
L0Provenance & inventoryWARN
L1Static analysis of the codePASS
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
declared (5 observation(s))
Network
none-observed
Shell
declared (3 observation(s))
Dependencies
pinned
Secrets in source
none-found

Findings (3)

MEDIUMInventory / provenance · inv.binary · CWE-1104
src/speech_mcp/resources/audio/start_listening.wav
start_listening.wav
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
MEDIUMInventory / provenance · inv.binary · CWE-1104
src/speech_mcp/resources/audio/stop_listening.wav
stop_listening.wav
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
LOWInventory / provenance · inv.hidden_file · CWE-1104
.goosehints
.goosehints
Why it matters. hidden member outside the usual dotfiles
Fix. review its purpose

Gates applied: no_behavioural_pass.

Audited 2026-10-07 · audit v0.4.1 · source sha dce2018e7b2bfull audit observations/trust-audit/mcp-server/kvadratni__speech.json · Report an issue / request a re-scan
05

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

DateSourceVerdictGradeScoreChange
2026-10-07dce2018e7b2bCAUTIONB89first audit
06

Questions

What is the Speech MCP server?

Speech MCP: A Goose MCP extension for voice interaction with audio visualization

What tools does Speech expose?

7 in total: 6 read-only, 1 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.

Is Speech safe to connect to an agent?

With care. The audit graded it B (89/100) and found 3 things worth knowing before you trust this server, listed below with the exact line each was found on.

What credentials does Speech need?

No credential environment variables were found in its source, so it appears to need none.

How current is this page?

The grade is for one exact copy of the source (dce2018e7b2b), read on 2026-10-07. The repository is watched and re-audited when it changes.

Advertisement