AI agents invoke run_tests to trigger actions in Neurodev. 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.
This tool executes arbitrary code (test suites) via pytest, making it an Execute category risk. The isolated environment provides some containment, limiting severity to medium rather than high. An agent could be tricked into running malicious test files that perform destructive or exfiltration actions outside the test assertions themselves.
From the tool's definition Tool description states 'Execute pytest tests' and 'Runs tests in an isolated environment' — the tool invokes an external testing framework (pytest) whose behavior depends on the test files present and their content.
Attacks that exploit this kind of access
Execute pytest tests with coverage reporting. Runs tests in an isolated environment and returns detailed results including pass/fail status, coverage percentage, and execution time. It is categorised as a Execute tool in the Neurodev MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Neurodev MCP server in PolicyLayer and add a rule for run_tests: 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 Neurodev. Nothing to install.
run_tests is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the run_tests 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 run_tests. 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.
run_tests is provided by the Neurodev MCP server (ravikant1918/neurodev-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 →