propose_distributions_for_inputs
AI agents call propose_distributions_for_inputs to retrieve information from ModelRisk MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Based on the tool name alone, this appears to propose or suggest probability distributions for model inputs, which is a read/analytical recommendation operation with no side effects. However, the empty description significantly lowers confidence. In context of sibling tools like 'list_distributions' and 'get_simulation_results', this likely returns recommendations rather than modifying anything.
From the tool's definition Tool name 'propose_distributions_for_inputs' suggests recommending/suggesting distributions for model inputs, which is a read/analytical operation. Description is empty and uninformative.
Attacks that exploit this kind of access
propose_distributions_for_inputs. It is categorised as a Read tool in the ModelRisk MCP MCP Server, which means it retrieves data without modifying state.
Register the ModelRisk MCP server in PolicyLayer and add a rule for propose_distributions_for_inputs: 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 ModelRisk MCP. Nothing to install.
propose_distributions_for_inputs 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 propose_distributions_for_inputs 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 propose_distributions_for_inputs. 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.
propose_distributions_for_inputs is provided by the ModelRisk MCP server (vosesoftware/modelrisk-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 →