Application Profiler Skill for Python cProfile Hotspots

Profile a local Python batch, identify repeated work by call count and cumulative time, and verify output parity after a change.

中文

Install in Claude Code

Use your own Python script, cProfile, and a fixed synthetic CSV fixture.

Get the complete `plugins/performance/application-profiler/skills/profiling-application-performance` directory from the pinned GitHub commit.
Copy it into the target project's `.claude/skills/`. Check `.claude/skills/profiling-application-performance/SKILL.md`, start Claude Code at the project root, and call `/profiling-application-performance`.

Example task

Run python -m cProfile -s cumulative summarize.py on 10,000 synthetic catalog records. Inspect repeated classification work, change one hotspot, and repeat with the same input.

What to check

The report points to observed call counts and cumulative time. Both output summaries remain comparable.

Original SKILL.md on GitHub · Updated 2026-11-13

Frequently asked question

Is cumulative time larger than self time a problem?

Cumulative time includes child calls. Use both measures to find where the work originates.