Sample from an LLM with multimodal content (text, images, audio)
AI agents call sampleMultimodal to retrieve information from MCP Elicitations Demo Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves or queries multimodal content from an LLM. It performs no side effects, does not create, modify, or delete data, does not execute arbitrary code, and does not involve financial transactions. The action is purely informational/retrieval in nature, consistent with the Read category.
From the tool's definition Tool description states 'Sample from an LLM with multimodal content (text, images, audio)' — a retrieval/query operation that fetches sampling results from an LLM without modifying any data.
Attacks that exploit this kind of access
Sample from an LLM with multimodal content (text, images, audio). It is categorised as a Read tool in the MCP Elicitations Demo Server MCP Server, which means it retrieves data without modifying state.
Register the MCP Elicitations Demo Server MCP server in PolicyLayer and add a rule for sampleMultimodal: 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 MCP Elicitations Demo Server. Nothing to install.
sampleMultimodal 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 sampleMultimodal 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 sampleMultimodal. 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.
sampleMultimodal is provided by the MCP Elicitations Demo Server MCP server (soriat/soria-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 →