Synthetic token efficiency benchmark: compare raw file reads vs trace-mcp compact responses across symbol lookup, file exploration, search, and impact analysis scenarios. Read-only, no side effects. Use to quantify token savings. Returns JSON: { scenarios: [{ name, raw_tokens, compact_tokens, sav...
AI agents call benchmark_project to retrieve information from Trace without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs benchmarking analysis by comparing token efficiency metrics across different lookup and search scenarios. It retrieves and compares data to generate performance statistics, with no capability to modify code, execute operations, or cause side effects. The explicit 'read-only' designation confirms it falls under the Read category with low severity risk.
From the tool's definition Tool description explicitly states 'Read-only, no side effects' and operations are limited to 'compare raw file reads vs trace-mcp compact responses' and 'Returns JSON' with metrics. No modification, execution, deletion, or financial operations occur.
Documented attack patterns abuse exactly the kind of access benchmark_project gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Trace, and nothing reaches the server without passing your rules. This is the rule we recommend for benchmark_project:
{
"version": "1",
"default": "deny",
"tools": {
"benchmark_project": {}
}
}benchmark_project 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.
Synthetic token efficiency benchmark: compare raw file reads vs trace-mcp compact responses across symbol lookup, file exploration, search, and impact analysis scenarios. Read-only, no side effects. Use to quantify token savings. Returns JSON: { scenarios: [{ name, raw_tokens, compact_tokens, savings_pct }], summary }. It is categorised as a Read tool in the Trace MCP Server, which means it retrieves data without modifying state.
Register the Trace MCP server in PolicyLayer and add a rule for benchmark_project: 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 Trace. Nothing to install.
benchmark_project 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 benchmark_project 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 benchmark_project. 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.
benchmark_project is provided by the Trace MCP server (nikolai-vysotskyi/trace-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 178 Trace tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
178 Trace tools catalogued and risk-classified — across an index of 42,500+ MCP servers.