Memory leak
Heap diffs, retention graphs, classic patterns (listeners, global caches)—applies across Node, browsers, JVM, etc.
Category · Troubleshooting & root cause
5 skills Category 20 of 20
This category goes deeper on hard-to-reproduce issues: heap snapshots for leaks, race conditions and ordering, concurrency backpressure, ADRs for architectural tradeoffs, and controlled codegen with review and tests. It complements observability—signals online vs. offline reasoning and durable knowledge.
In the hub it is the final category—compare with entries in the case hub. The five skills match the main hub.
Heap diffs, retention graphs, classic patterns (listeners, global caches)—applies across Node, browsers, JVM, etc.
Sequence diagrams for interleavings, locks or lock-free structures to remove data races—often overlaps flaky tests.
Worker pools, bounded queues, backpressure, timeouts—graceful degradation under overload instead of unbounded buffers.
Capture context, options, and consequences so future readers know “why it is this way”—pairs with RFC-style docs at different granularity.
Treat OpenAPI/Proto-generated code as read-only in review and CI; template changes need snapshot tests.