profiling▌

Iteratively optimize cuTile kernel performance through systematic profiling, bottleneck analysis, IR comparison, and targeted tuning. Covers tile sizes, occupancy, autotune configs, TMA, latency hints, persistent scheduling, num_ctas, flush_to_zero, and IR-level debugging. Use when asked to "optimize cutile kernel", "improve kernel perf", "tune cutile performance", "make kernel faster", or iteratively benchmark and refine a cuTile GPU kernel in the TileGym project.

Top-level workflow skill for USD performance diagnosis and optimization. Use for slow loading, high memory, low FPS, or 'optimize my scene' requests; delegates auth/runtime setup to Phase 0 owners.

Build comprehensive threat actor profiles using open-source intelligence (OSINT) techniques to document adversary motivations, capabilities, infrastructure, and TTPs for proactive defense.

Extract cognitive patterns and thinking fingerprints from any text to analyze reasoning and decision-making styles.

Optimizes application performance by identifying and fixing bottlenecks based on profiling data.

Profile code execution to identify performance bottlenecks and optimize critical paths using data-driven approaches.

Profile and optimize Swift async/await code using Instruments.

Systematic approach to measuring, analyzing, and optimizing application performance. \n \n Covers Core Web Vitals (LCP, INP, CLS) with target thresholds and measurement stages from development through production \n Provides a 4-step profiling workflow: establish baseline, identify bottleneck, apply fix, validate improvement \n Includes tool selection guidance for specific problems (Lighthouse for page load, DevTools for runtime and memory, bundle analyzers for code size) \n Documents common runt

Comprehensive statistical and quality analysis of database tables with structured profiling output. \n \n Generates column-level statistics tailored to data type: min/max/percentiles for numeric columns, length metrics for strings, date ranges for timestamps \n Performs cardinality analysis to identify categorical vs. high-cardinality columns and detect skewed distributions \n Assesses data quality across five dimensions: completeness (NULL rates), uniqueness (duplicates), freshness (update time

CPU profiling identifies which functions consume most CPU time, enabling targeted optimization of expensive code paths.