AI agents call ground_probability to retrieve information from Musashi without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves market probability data to inform the user, with no side effects, modifications, or irreversible actions. It is purely informational and falls under the Read category (query/fetch). Severity is low since misuse would only return potentially misleading market information without causing direct harm, data loss, or financial commitment.
From the tool's definition Tool description indicates it 'grounds a real-world claim in live market-implied probability'—a data retrieval operation that queries market data without modifying, executing external code, or committing financial resources.
Attacks that exploit this kind of access
Ground a real-world claim in live market-implied probability. Use this when the user wants to know the market\. It is categorised as a Read tool in the Musashi MCP Server, which means it retrieves data without modifying state.
Register the Musashi MCP server in PolicyLayer and add a rule for ground_probability: 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 Musashi. Nothing to install.
ground_probability 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 ground_probability 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 ground_probability. 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.
ground_probability is provided by the Musashi MCP server (musashibot/musashi-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 →