get_trajectory_analysis
Analyze tool usage trajectories across sessions. Returns tool frequency, error rates, average duration, phase distribution, and sequential patterns. Use this to understand how tools are actually being used and identify bottlenecks.
This record as markdown: /tools/io-github-homenshum-nodebench/get-trajectory-analysis.md
What get_trajectory_analysis does on Nodebench
AI agents call get_trajectory_analysis to retrieve information from Nodebench without modifying anything. It is typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Why get_trajectory_analysis is rated Low
This tool retrieves and analyzes historical usage data to generate insights. It performs read-only operations that examine existing data without creating, modifying, deleting, or executing external actions. The blast radius of misuse is minimal—worst case would be disclosure of usage analytics which, while potentially sensitive, is passive introspection rather than system-altering action.
From the tool's definition Tool description states it 'Analyze[s]' and 'Returns' data about tool usage trajectories, frequencies, error rates, durations, and patterns. These are query/retrieval operations with no modification or execution of external systems.
Attacks that exploit this kind of access
The rule that runs get_trajectory_analysis safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Nodebench, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For get_trajectory_analysis, this is the rule to start with:
get_trajectory_analysis is read-only, so it stays allowed. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Nodebench, apply this rule, and every get_trajectory_analysis call is checked against it from then on.
Questions about get_trajectory_analysis
Analyze tool usage trajectories across sessions. Returns tool frequency, error rates, average duration, phase distribution, and sequential patterns. Use this to understand how tools are actually being used and identify bottlenecks. It is categorised as a Read tool in the Nodebench MCP Server, which means it retrieves data without modifying state.
Register the Nodebench MCP server in PolicyLayer and add a rule for get_trajectory_analysis: 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 Nodebench. Nothing to install.
get_trajectory_analysis 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_trajectory_analysis 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_trajectory_analysis. 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_trajectory_analysis is provided by the Nodebench MCP server (nodebench-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Nodebench, and thousands of servers like it.
This server
Across the catalogue