performance-tuning
Skill8 starsChanged 2 months agoMITView on GitHub
Use when the request is a performance symptom — slow endpoint, high latency, timeouts, high CPU or memory, big bundle, slow build. Measurement-first methodology for the Hydraia perf route: baseline before analysis, numeric target in the spec, re-measure in verify. Never optimize without measuring.
8 other projects use a version of this.
Adopted by
Most recently taken up by
- NoobyGains/godmodetheir own version, first seen Nov 1, 2026
- 0King0P/Claude-Skillstheir own version, first seen Oct 13, 2026
- devpilotX/skillstheir own version, first seen Sep 22, 2026
- kiko197/skilltheir own version, first seen Aug 11, 2026
- SpencerGoss/agent-engineeringtheir own version, first seen Jun 18, 2026
- polardb/polardb-pg-skillstheir own version, first seen May 11, 2026
Used with
Projects with it also use code-review.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
GoogleSkillcode-review
Adversarial systems, concurrency, memory-model, and architecture code review protocol. Use this skill whenever asked to review code, PRs, diffs, or branches in Filament.
Also in 4 projects that use this
Other versions8 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 8
- NoobyGains/godmode★ 118same name
- polardb/polardb-pg-skills★ 2same name
- harrissohail6/claude-code-config★ 1same name
- 0King0P/Claude-Skills★ 0same name
- devpilotX/skills★ 0same name
- kiko197/skill★ 0same name
- poteto0/my-skills★ 0same name
- SpencerGoss/agent-engineering★ 0same name
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 repository51 other files it gives its agents.
- adrSkill
- agency-tier-uiSkill
- api-designSkill
- banner-designSkill
- brainstormingSkill
- brandSkill
- cavemanSkill
- coding-standardsSkill
- db-optimizationSkill
- designSkill
- design-systemSkill
- design-to-codeSkill
- dispatching-parallel-agentsSkill
- django-securitySkill
- dotnet-patternsSkill
- e2e-testingSkill
- executing-plansSkill
- explainmeSkill
- finishing-a-development-branchSkill
- golang-patternsSkill
- greenfield-architectSkill
- hydraiaSkill
- karpathy-guidelinesSkill
- microservices-architectSkill
- mobile-design-framesSkill
- node-patternsSkill
- production-auditSkill
- python-patternsSkill
- react-patternsSkill
- receiving-code-reviewSkill
- redesign-upliftSkill
- repo-scanSkill
- requesting-code-reviewSkill
- security-bounty-hunterSkill
- security-reviewSkill
- security-scanSkill
- slidesSkill
- springboot-patternsSkill
- springboot-securitySkill
- story-analysisSkill
- subagent-driven-developmentSkill
- systematic-debuggingSkill
- test-driven-developmentSkill
- ui-stylingSkill
- ui-ux-pro-maxSkill
- using-git-worktreesSkill
- using-superpowersSkill
- verification-before-completionSkill
- web-design-framesSkill
- writing-plansSkill
- writing-skillsSkill
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 11
NO optimization without a measured baseline.
- Distinctive: line 21
Baseline: Run the measurement on representative data (state dev-vs-prod caveats).
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Performance Tuning — measure first, always
The performance route through the Hydraia pipeline. The pipeline phases are unchanged; this skill defines the two contract adaptations and the hard rules.
<HARD-RULES>
- NO optimization without a measured baseline. If there is no number, the first deliverable is the measurement, not the fix.
- The Phase 2 spec MUST contain: the baseline (metric, command, environment, runs, spread) and a NUMERIC target. A perf spec without numbers is not a spec.
- One variable at a time: each plan task isolates one change so Phase 6 re-measurement can attribute the gain.
- Phase 6 verify RE-MEASURES with the same commands as the baseline and compares against the target. Target missed → the run is NOT done: one more hypothesis cycle (max one), or an honest shortfall report. Never redefine the target post-hoc.
- Rejected hypotheses stay in the spec — what you chose NOT to optimize and why is part of the record.
</HARD-RULES>Route flow
- Symptom → metric. Express the complaint as a measurable number (p50/p95 latency, throughput, CPU %, RSS, bundle KB, query ms).
- Baseline. Run the measurement on representative data (state dev-vs-prod caveats). No access → the human runs the command you provide; the pipeline waits (interactive half).
- Diagnose. Dispatch
perf-engineerwith the symptom + baseline. DB-shaped findings → it delegates todb-performance-tuner(dispatch that too). Their reports feed Phase 2. - Spec. Baseline + target + ranked hypotheses (chosen AND rejected) + threat model as usual. Then plan, freeze, gate — the normal pipeline from here.
- Execute. Sonnet executors implement; one isolated change per task.
- Verify. Re-run the exact baseline commands. Compare. Report measured before/after per change. Close only at/beyond target, or after the one allowed iteration with the shortfall stated plainly.
Measurement pointers (per stack — pointers, not tutorials)
- Node:
node --cpu-prof,clinic flame,0x; web: Chrome DevTools tracing, Lighthouse, Web Vitals. - Python:
cProfile+snakeviz,py-spy record(live processes). - Go:
go test -bench,pprof(CPU/heap),trace. - JVM:
async-profiler, JFR. - CLI/anything:
hyperfinefor wall-clock A/B. - DB: see the db-optimization skill.
Report medians and percentiles over multiple runs, never a single run. State machine load conditions.
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.