AI agents invoke start-ec2-instance to trigger actions in AWS MCP Server. 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 a command that starts an EC2 instance, which is an external operation with real-world consequences. While not strictly destructive (the instance already exists and is being started, not deleted), it is Execute because it triggers an operation whose effects cannot be easily reversed and depend entirely on which instance is targeted.
From the tool's definition start-ec2-instance: Start an EC2 instance in a given region. This tool triggers external operations (EC2 instance startup) whose effects depend on arguments (region, instance ID). The action is irreversible in practical terms and consumes resources/costs.
Documented attack patterns abuse exactly the kind of access start-ec2-instance gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and AWS MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for start-ec2-instance:
{
"version": "1",
"default": "deny",
"tools": {
"start-ec2-instance": {
"limits": [
{
"counter": "start-ec2-instance_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}start-ec2-instance stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Start an EC2 instance in a given region. It is categorised as a Execute tool in the AWS MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the AWS MCP Server MCP server in PolicyLayer and add a rule for start-ec2-instance: 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 AWS MCP Server. Nothing to install.
start-ec2-instance 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 start-ec2-instance 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 start-ec2-instance. 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.
start-ec2-instance is provided by the AWS MCP Server MCP server (lokeswaran-aj/aws-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from AWS 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.
54 AWS MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.