AI agents use send_metric to create or update resources in Otel — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Otel environment.
Sending metrics is a write operation that creates new observability data in a telemetry system. While reversible (metrics can be cleared or overwritten), it modifies the state of the monitoring/logging infrastructure. The severity is medium because misuse could pollute telemetry data, create false alerts, or consume resources, but does not directly delete data, execute arbitrary code, or move money.
From the tool's definition Tool name 'send_metric' and description 'Send one or more metric instruments' indicate the tool creates or modifies telemetry data by writing metrics to OTLP endpoints.
Attacks that exploit this kind of access
Send one or more metric instruments. It is categorised as a Write tool in the Otel MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Otel MCP server in PolicyLayer and add a rule for send_metric: 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 Otel. Nothing to install.
send_metric 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 send_metric 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 send_metric. 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.
send_metric is provided by the Otel MCP server (probsjustin/otel_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 →