Transcribe audio to text using Voxtral STT. Supported languages: ${languageCodes}
AI agents call stt_transcribe to retrieve information from Mistral Ai without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Speech-to-text transcription is a read-only operation that processes input audio and returns text output. It does not modify, delete, or execute code; it does not move money; and it does not trigger external system changes.
From the tool's definition Tool description states 'Transcribe audio to text' — a one-way conversion that retrieves/extracts text content from audio input with no side effects, creation, modification, deletion, or execution of external operations.
Attacks that exploit this kind of access
Transcribe audio to text using Voxtral STT. Supported languages: ${languageCodes}. It is categorised as a Read tool in the Mistral Ai MCP Server, which means it retrieves data without modifying state.
Register the Mistral Ai MCP server in PolicyLayer and add a rule for stt_transcribe: 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 Mistral Ai. Nothing to install.
stt_transcribe 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 stt_transcribe 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 stt_transcribe. 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.
stt_transcribe is provided by the Mistral Ai MCP server (tan-yong-sheng/mistral-ai-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 →