AI agents use amplitude_track_event to create or update resources in Amplitude MCP Server — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Amplitude MCP Server environment.
This tool creates new analytics event records in Amplitude, modifying the analytics dataset irreversibly in practice (events cannot meaningfully be 'untracked'). However, it is classified as Write rather than Destructive because: (1) the primary intent is data creation/logging, not deletion; (2) individual events are relatively low-impact compared to bulk deletion; (3) the operation is a standard analytics function.
From the tool's definition Tool description states 'Track a custom event in Amplitude' and is part of a server that 'enables AI assistants...to track events'.
Documented attack patterns abuse exactly the kind of access amplitude_track_event gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Amplitude MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for amplitude_track_event:
{
"version": "1",
"default": "deny",
"tools": {
"amplitude_track_event": {
"limits": [
{
"counter": "amplitude_track_event_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}amplitude_track_event 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.
Track a custom event in Amplitude. It is categorised as a Write tool in the Amplitude MCP Server MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Amplitude MCP Server MCP server in PolicyLayer and add a rule for amplitude_track_event: 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 Amplitude MCP Server. Nothing to install.
amplitude_track_event 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 amplitude_track_event 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 amplitude_track_event. 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.
amplitude_track_event is provided by the Amplitude MCP Server MCP server (moonbirdai/amplitude-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Amplitude 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.
5 Amplitude MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.