Listen to microphone and transcribe
AI agents call voice_listen to retrieve information from Voice-AGI MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool performs audio input capture and transcription, which is a read-only operation that retrieves and transforms user-provided voice data into text. It has no capability to modify, delete, execute code, or commit financial transactions. The severity is low because misuse would only affect data retrieval without cascading harmful consequences.
From the tool's definition Tool description states 'Listen to microphone and transcribe' — this retrieves audio data from the microphone and converts it to text. No side effects, data modification, code execution, or irreversible actions occur.
Attacks that exploit this kind of access
Listen to microphone and transcribe. It is categorised as a Read tool in the Voice-AGI MCP Server MCP Server, which means it retrieves data without modifying state.
Register the Voice-AGI MCP Server MCP server in PolicyLayer and add a rule for voice_listen: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Voice-AGI MCP Server. Nothing to install.
voice_listen is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the voice_listen rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for voice_listen. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
voice_listen is provided by the Voice-AGI MCP Server MCP server (marc-shade/voice-agi-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →