AI agents use km_add_action to create or update resources in Keyboard Maestro — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Keyboard Maestro environment.
This tool modifies an existing macro by adding a new action to it, which is a write/create operation. However, since the action being added could be anything (including shell scripts, destructive operations, etc.), the severity is high. The tool itself is a Write operation — it modifies macro definitions — but the content of what's added could be more dangerous.
From the tool's definition Add an action to an existing macro using XML
Documented attack patterns abuse exactly the kind of access km_add_action gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Keyboard Maestro, and nothing reaches the server without passing your rules. This is the rule we recommend for km_add_action:
{
"version": "1",
"default": "deny",
"tools": {
"km_add_action": {
"limits": [
{
"counter": "km_add_action_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}km_add_action stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Add an action to an existing macro using XML. It is categorised as a Write tool in the Keyboard Maestro MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Keyboard Maestro MCP server in PolicyLayer and add a rule for km_add_action: 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 Keyboard Maestro. Nothing to install.
km_add_action is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the km_add_action 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 km_add_action. 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.
km_add_action is provided by the Keyboard Maestro MCP server (saihgupr/keyboard-maestro-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Keyboard Maestro, 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.
34 Keyboard Maestro tools catalogued and risk-classified — across an index of 43,000+ MCP servers.