run_pytest

What run_pytest does on Mapi Agent Memory

AI agents invoke run_pytest to trigger actions in Mapi Agent Memory. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.

Why run_pytest is rated High

pytest execution can trigger arbitrary test code, modify repository state (create temp files, fixtures with side effects), and potentially interact with external systems depending on test definitions. While not inherently destructive, the ability to execute code in a controlled test environment still represents a significant risk if an AI agent invokes it with malicious or unintended test selections.

From the tool's definition Tool description states 'Run pytest from the repository root' — this directly executes code (pytest test runner) whose effects depend on which tests are selected and the state of the repository.

Risk signalsAdmin/system-level operation

The rule that runs run_pytest safely

PolicyLayer is an MCP gateway: it sits between your AI agents and Mapi Agent Memory, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For run_pytest, this is the rule to start with:

{
  "version": "1",
  "default": "deny",
  "tools": {
    "run_pytest": {
      "limits": [
        {
          "counter": "run_pytest_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

Questions about run_pytest