hammer-memory-profiler

Memory leak detection uses Valgrind memcheck and AddressSanitizer to identify unreleased allocations, invalid reads/writes, use-after-free errors, and double-frees in your SDL3 HammerEngine code. Run a Quick Leak Check or Full Memory Profile to scan your codebase and generate a detailed report with specific leak locations and counts.

Allocation Profiling identifies which subsystems (AI, Collision/Pathfinding, Core, Resources) allocate memory each frame and where buffer reuse anti-patterns occur. The profiler generates per-system breakdowns and optimization recommendations to eliminate performance spikes and frame drops.

Target-specific subsystem profiling lets you isolate memory behavior for AI, Collision/Pathfinding, Core, and Resources independently. Use modes like Quick Leak Check, Allocation Profiling, or Full Memory Profile to analyze each system separately, compare baselines over time, and track optimization progress.

Yes. This Skill integrates both Valgrind memcheck and massif with AddressSanitizer to provide layered error detection and allocation profiling. Valgrind catches leaks and memory errors; AddressSanitizer detects runtime violations; together they give comprehensive SDL3 HammerEngine memory visibility.

Run a Full Memory Profile after implementing changes, then compare the baseline against previous reports. The Skill generates Markdown reports with statistical analysis and baseline comparisons, letting you catch new leaks, per-frame allocations, and buffer reuse violations before they degrade performance.

Memory profiling requires Valgrind, Python3, CMake, and Ninja. The Skill ships with Bash scripts and a Python massif parser for automated analysis, so setup is automated once dependencies are installed. Verify all four tools are available before running profiles.