performance-engineer
Skill1 starsChanged 3 months agoMITView on GitHub
Evidence-based performance optimization: benchmark the real code before and after changes, profile to find the true bottleneck, patch, and verify improvement with statistical rigor. Use when code is slow, latency needs reducing, or optimization claims must be proven.
14 other projects use a version of this.
Adopted by
Most recently taken up by
- xfstudio/skillstheir own version, first seen Oct 13, 2026
- Tgoldi/claude-skillstheir own version, first seen Oct 13, 2026
- Kshitijpalsinghtomar/depth-skillstheir own version, first seen Oct 1, 2026
- rpkchtrj/aether-workflowtheir own version, first seen Sep 27, 2026
- suhanr/next-gen-claude-skillstheir own version, first seen Aug 21, 2026
- cdgutierrez6/claude-skillstheir own version, first seen Aug 21, 2026
Other versions14 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 14
- jshsakura/awesome-opencode-skills★ 30same name
- belokonm/claude-supercode-skills★ 11same name
- rade-nugroho/Agentic-Enteprises-Skill★ 9same name
- xfstudio/skills★ 9same name
- daemon-blockint-tech/Agentic-Enteprises-Skill★ 8same name
- itsthatladydev/my-agent-skills★ 6same name
- Kshitijpalsinghtomar/depth-skills★ 5same name
- ai4brands-design/claude-skills★ 2same name
- suhanr/next-gen-claude-skills★ 2same name
- cdgutierrez6/claude-skills★ 0same name
4 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- codebase-architectSkill
- deep-debuggerSkill
- dependency-doctorSkill
- documentation-engineerSkill
- refactor-engineerSkill
- release-engineerSkill
- repo-resurrectionSkill
- security-auditorSkill
- test-engineerSkill
The file
Files
Performance Engineer
Tối ưu dựa trên số liệu THẬT, không dựa trên cảm giác. Pipeline: benchmark trước → profile → xác định bottleneck → patch → benchmark lại → so sánh thống kê.
Quy trình (6 bước)
- Benchmark baseline: chạy code thật nhiều lần (≥5 iterations), lấy median + p95. Ghi environment (OS, Python, CPU). Không dùng 1 lần đo.
- Profile: dùng cProfile/timeit/line_profiler trên workload thật. Xác định top-3 hàm tốn thời gian nhất.
- Identify bottleneck: chọn 1 bottleneck lớn nhất có evidence rõ (phần trăm thời gian). KHÔNG tối ưu nhiều chỗ cùng lúc.
- Patch: sửa tối thiểu tại bottleneck. Ghi rõ giả thuyết cải thiện.
- Benchmark again: cùng workload, cùng environment, ≥5 iterations. Tính speedup = median_before / median_after.
- Verify: speedup ≥20% mới coi là significant; kiểm tra correctness (output không đổi); kiểm tra không phá trade-off (memory, readability).
Nguyên tắc
- Median, không mean (chống outlier).
- Warmup trước khi đo.
- Một thay đổi một lần — isolate variable.
- Không optimze premature: chỉ tối ưu khi có số liệu chậm thật.
- Nếu patch làm chậm: revert, ghi lại learning.
Scripts
scripts/profile_code.py <target> [--iterations N]— wrapper cProfile + timing median.scripts/compare.py before.json after.json— so 2 kết quả benchmark, tính speedup + verdict.
References
- PROFILING-TOOLS.md
- STATS-NOTES.md — tại sao median, outlier handling
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.