AI agents call search_cases to retrieve information from Legal without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read-only operation that searches and retrieves existing legal case data. The description explicitly states it 'Search[es]' and 'Returns' data with no mention of modification, creation, deletion, or execution capabilities. All sibling tools (get_case, read_opinion, list_* functions) are also read operations. There is no side effect or state change—it simply queries the legal database.
From the tool's definition Tool name 'search_cases' and description 'Search US case law (court opinions). Returns normalized' indicate a query/search operation that retrieves data from the CourtListener API without modifying, executing, or deleting anything.
Attacks that exploit this kind of access
Search US case law (court opinions). Returns normalized. It is categorised as a Read tool in the Legal MCP Server, which means it retrieves data without modifying state.
Register the Legal MCP server in PolicyLayer and add a rule for search_cases: 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 Legal. Nothing to install.
search_cases 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 search_cases 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 search_cases. 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.
search_cases is provided by the Legal MCP server (mcpservings/legal-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 →