stop_profiling
June 8, 2026 ยท View on GitHub
Stop performance measurement on the active debug target. Counterpart to start_profiling: deterministically switches profiling off. Idempotent: if profiling is not active for this applicationId it returns a benign result, not an error. Call get_profiling_results afterwards to retrieve the collected coverage.
Parameters
| Parameter | Required | Type | Description |
|---|---|---|---|
| applicationId | yes | string | Application id of the running debug session (required) |
Guide
Turns off the line-level profiling started by start_profiling for a debug session. Stop it once your scenario has run, then call get_profiling_results to retrieve the collected coverage and timing.
When to use
- Your scenario has finished and you want to end measurement before reading results.
- To reset profiling state for an application before starting a fresh measurement.
Parameter details
applicationId(required) - the same debug session you passed tostart_profiling.
What you get
JSON: active (false after a successful stop), stopped (whether it was actually toggled off), applicationId, and a message.
Notes & gotchas
- Idempotent: if profiling was never started (or already stopped) for this id you get a benign success (
stopped: false), not an error. - If the debug session has already ended, the tool clears its tracked state and reports that benignly - your collected results are still retrievable.
- Stopping does not by itself print the data - follow with
get_profiling_results.
Generated from the live MCP server (get_tool_guide) by docs/generate_tool_docs.py. Do not edit this file. Edit the tool's description/schema in its Java source and its guide body in mcp/bundles/com.ditrix.edt.mcp.server/guides/<tool>.md.