Compares multiple test runs to identify performance changes.
AI agents call compare_test_runs to retrieve information from MockLoop MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool reads and analyzes existing test run data to perform comparisons and identify trends. It retrieves information for reporting purposes with no capability to modify test runs, execute new operations, or trigger side effects. The action is purely analytical and non-destructive, placing it firmly in the Read category with low severity.
From the tool's definition Tool name 'compare_test_runs' and description 'Compares multiple test runs to identify performance changes' indicate data retrieval and analysis operations without modification or execution of external systems.
Documented attack patterns abuse exactly the kind of access compare_test_runs gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and MockLoop MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for compare_test_runs:
{
"version": "1",
"default": "deny",
"tools": {
"compare_test_runs": {}
}
}compare_test_runs is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Compares multiple test runs to identify performance changes. It is categorised as a Read tool in the MockLoop MCP Server MCP Server, which means it retrieves data without modifying state.
Register the MockLoop MCP Server MCP server in PolicyLayer and add a rule for compare_test_runs: 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 MockLoop MCP Server. Nothing to install.
compare_test_runs 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 compare_test_runs 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 compare_test_runs. 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.
compare_test_runs is provided by the MockLoop MCP Server MCP server (mockloop/mockloop-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from MockLoop MCP Server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
30 MockLoop MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.