matlab-profilelisted

Investigate MATLAB performance problems: time hotspots, memory consumption, parallelism, call patterns. Configures the profiler, runs code under instrumentation, analyzes results across multiple passes, and reports findings with automatic before/after comparison. Use this skill when the user says "profile", "matlab profile", "why is this slow", "memory usage", "hotspot", "performance", "where is time going", "bottleneck", or invokes /matlab-profile. Also use when the user asks to optimize MATLAB code and you need to measure before changing.
richardkmichael/claude-rodin · ★ 0 · Data & Documents · score 75

Install: claude install-skill richardkmichael/claude-rodin

# MATLAB Performance Investigation Investigate performance problems in the user's MATLAB code. The profiler is a diagnostic tool — choose the right options and passes based on what you are trying to learn. Iterate to build a complete picture, then report findings. Do the investigation; the user decides what to change. Read `${CLAUDE_SKILL_DIR}/references/PROFILER_REFERENCE.md` at the start of every investigation for profiler options, output struct schema, and interpretation rules. ## Tools All scripts are in `${CLAUDE_SKILL_DIR}/scripts/`. Invoke them directly; they are executable. Analysis (take `profile.json`): - `rank_functions.py <profile.json> [--top N] [--by time|memory|calls]` - `line_breakdown.py <profile.json> [<function>] [--top N] [--lines M]` - `compare_runs.py <before.json> <after.json> [--top N]` MATLAB-side helpers (inline into `matlab -batch` commands or run via `matlab -batch "run('...'); ..."`): - `export_profile_json.m` — function `export_profile_json(info, outPath)`, serialises `profile('info')` to JSON. - `history_query.m` — function `events = history_query(matPath[, fnNames])`, flattens FunctionHistory for call-pattern analysis. Documentation lookup: use the `matlab-docs` skill when an unfamiliar function appears in results. ## Constraints - Do NOT optimize or refactor the user's code. The skill profiles and reports; the user decides what to change. Temporary instrumentation (profiler guards, tic/toc, input-capture saves for snapshot-and