amayo/.github/prompts/resource-checker.prompt.md

Prompt: Node.js Resource Checker (Memory & CPU)

Whenever Copilot generates, modifies, or reviews code in Node.js:

1. Memory Monitoring

   • Always remind to check memory usage with:

     logger.info(process.memoryUsage());
     

     which returns RSS, Heap Used, Heap Total, External, and Array Buffers.
   • If Heap Used grows continuously without freeing, suggest potential memory leaks (unreleased timers, listeners, or large cached data).
   • If RSS (Resident Set Size) is unusually higher than Heap Total, warn that external/native modules might be consuming memory.

2. CPU Usage

   • Remind to check CPU load using:

     top -p <pid>
     

     or Node.js profiling (--inspect, clinic flame).
   • Flag infinite loops, synchronous blocking calls, or heavy computations running on the main thread as potential CPU overuse.
   • Encourage offloading heavy tasks to Workers, child processes, or queues when appropriate.

3. Best Practices

   • Always validate memory and CPU usage after implementing new features.
   • Suggest garbage collection checks (global.gc()) in development when Node is run with --expose-gc.
   • Warn if long-living references (e.g., large Maps, Sets, caches, or listeners) might not be released.
   • Highlight that CPU and memory efficiency are as important as correctness.

4. Output Analysis

   • When given memory stats like:

     RSS: 186.9MB
     Heap Used: 67.6MB
     Heap Total: 71.2MB
     External: 5.0MB
     

     • Compare Heap Used vs Heap Total: if close to the limit, risk of OOM.
     • Compare RSS vs Heap Total: if RSS is much larger, check for native module or buffer leaks.
     • If growth is unbounded, warn about potential memory leaks.

5. Always remind to rerun tests with npx tsc --noEmit (for type safety) and memory checks together, ensuring both correctness and performance.