AI agents call trace_step to retrieve information from Novyx without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The name implies querying or retrieving trace/debugging information about a step in execution history, which is consistent with Read operations (introspection, audit). Given the server's focus on audit and governance, a trace function likely retrieves execution details without modifying state. However, the empty description reduces confidence.
From the tool's definition Tool name 'trace_step' suggests retrieval or inspection of execution steps or trace information. Description is empty, limiting definitive classification.
Documented attack patterns abuse exactly the kind of access trace_step gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Novyx, and nothing reaches the server without passing your rules. This is the rule we recommend for trace_step:
{
"version": "1",
"default": "deny",
"tools": {
"trace_step": {}
}
}trace_step 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.
trace_step. It is categorised as a Read tool in the Novyx MCP Server, which means it retrieves data without modifying state.
Register the Novyx MCP server in PolicyLayer and add a rule for trace_step: 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 Novyx. Nothing to install.
trace_step 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 trace_step 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 trace_step. 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.
trace_step is provided by the Novyx MCP server (novyxlabs/novyx-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 120 Novyx tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
120 Novyx tools catalogued and risk-classified — across an index of 42,500+ MCP servers.