Rank cross-module file edges by how unexpected they look (deep folder distance + popular target + few edges = high surprise). Surfaces hidden coupling that shotgun-changes through unrelated modules. Requires detect_communities to have been run first. Read-only. Returns JSON: { edges: [{ sourceFil...
AI agents call get_surprises 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.
get_surprises retrieves and ranks pre-computed cross-module dependency edges to identify unexpected coupling patterns. It performs data analysis and ranking on an existing dependency graph (requires detect_communities to have run first) and returns structured results. No code execution, modification, deletion, or financial operations occur.
From the tool's definition Tool description explicitly states 'Read-only' and 'Returns JSON'. It ranks and surfaces existing dependency information without modifying code, executing operations, or triggering external effects. The operation is purely analytical.
Documented attack patterns abuse exactly the kind of access get_surprises 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 get_surprises:
{
"version": "1",
"default": "deny",
"tools": {
"get_surprises": {}
}
}get_surprises 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.
Rank cross-module file edges by how unexpected they look (deep folder distance + popular target + few edges = high surprise). Surfaces hidden coupling that shotgun-changes through unrelated modules. Requires detect_communities to have been run first. Read-only. Returns JSON: { edges: [{ sourceFile, targetFile, surpriseScore, ... }], totalCommunities }. 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 get_surprises: 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.
get_surprises 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_surprises 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_surprises. 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_surprises 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.