Currently, perf try jobs where specific tests are run can be started for Telemetry tests using run_benchmark, and perf try jobs will be run on the CQ when a CL includes changes to a benchmark. However, sometimes someone might suspect that a patch has performance consequences, but is uncertain which benchmarks or metrics it might affect.

Annie, Ned, is there a way to do this now? Is this something we'd like to add in the future?

Note: This is related to an older feature request:  issue 324814 .