AI agents call select_match to retrieve information from Mr Magic without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool reads/selects from a prior search result set using filters. It does not create, modify, delete, or execute anything — it simply picks one item from existing results. Severity is low as misuse would only affect which result is selected, not any external system.
From the tool's definition 'Pick a single match from a previous search result based on provider/index filters' — this selects/filters from already-retrieved data with no indication of side effects
Attacks that exploit this kind of access
Pick a single match from a previous search result based on provider/index filters. Send params.arguments as an object (not a JSON string) for multiline payload safety. It is categorised as a Read tool in the Mr Magic MCP Server, which means it retrieves data without modifying state.
Register the Mr Magic MCP server in PolicyLayer and add a rule for select_match: 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 Mr Magic. Nothing to install.
select_match 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 select_match 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 select_match. 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.
select_match is provided by the Mr Magic MCP server (mr-magic-mcp-server). 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 →