AI agents call get-trace to retrieve information from Phoenix without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The 'get-' prefix strongly indicates a retrieval operation with no modification or execution. Given the context of sibling tools that include read operations (get-dataset, get-dataset-examples, get-dataset-experiments, get-experiment-by-id) and analytical operations, this tool most likely retrieves trace/log data for inspection purposes.
From the tool's definition Tool name 'get-trace' suggests retrieval of trace data (logs, execution traces, or diagnostic information). The naming convention 'get-*' is consistent with read-only query operations. No description provided.
Documented attack patterns abuse exactly the kind of access get-trace gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Phoenix, and nothing reaches the server without passing your rules. This is the rule we recommend for get-trace:
{
"version": "1",
"default": "deny",
"tools": {
"get-trace": {}
}
}get-trace is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
get-trace. It is categorised as a Read tool in the Phoenix MCP Server, which means it retrieves data without modifying state.
Register the Phoenix MCP server in PolicyLayer and add a rule for get-trace: 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 Phoenix. Nothing to install.
get-trace is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the get-trace 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 get-trace. 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.
get-trace is provided by the Phoenix MCP server (@Arize-ai/phoenix). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 34 Phoenix tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
34 Phoenix tools catalogued and risk-classified — across an index of 42,500+ MCP servers.