openclaw-test-heap-leaks

Compare two heap snapshot files or a lane directory with matching snapshots to surface prioritized delta summaries of dominant retained objects. This directly surfaces memory growth and retention hotspots to guide targeted fixes for OOM failures in OpenClaw tests.

Heap delta analysis identifies memory growth by comparing two heap snapshots to surface deltas and retention hotspots. It outputs a prioritized summary of dominant retained objects, reducing guesswork in OOM failures by highlighting specific per-object-type memory growth.

Apply heap delta analysis to CI lanes with multiple PIDs and repeated snapshots to surface differences across workers. By specifying a lane directory with matching snapshots, you get a prioritized delta summary highlighting dominant retained objects across different worker processes.

Yes, this analysis works with lane-based CI captures like unit-fast and unit-fast-batch-*. It compares matching heap snapshots from these directories to surface memory deltas and dominant retained objects, triaging memory growth in tests across multiple workers.

Specify two heap snapshot files or a lane directory with matching snapshots. The analysis compares these inputs to output a prioritized delta summary of dominant retained objects and actionable next steps for targeted fixes.