Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
affaan-m avatar

Benchmark Optimization Loop

  • 1.3k installs
  • 238k repo stars
  • Updated August 5, 2026
  • affaan-m/ecc

benchmark-optimization-loop is a Claude Code skill that converts vague speed requests into bounded, measured optimization loops for developers who need latency, throughput, or cost improvements with correctness preserved

About

benchmark-optimization-loop is an ECC-origin Claude Code skill that replaces open-ended requests like make it 20x faster with structured, measured optimization cycles. Before any change, the skill requires a defined operation, a correctness gate that must stay green, and a chosen metric such as wall time, p95 latency, rows per second, cost per run, or memory usage. Developers reach for benchmark-optimization-loop when comparing implementation variants, running recursive optimization attempts, or choosing the best approach through repeated measured tests rather than intuition. The skill uses Read, Write, Edit, Bash, Grep, and Glob tools to iterate on code while holding benchmarks accountable. It fits backend services, data pipelines, and CLI tools where performance regressions are costly.

  • Converts "make it 20x faster" or "try 50 recursive optimizations" into a measurable loop
  • Requires baseline, correctness gate, chosen metric, and search budget before starting
  • 8-step loop: measure baseline, identify bottlenecks, generate hypothesis-driven variants, run tests, reject unsafe ones,
  • Maintains a Variant Table tracking hypothesis, command, time, correctness, and notes
  • Hard gate: never optimize without existing operation, correctness test, metric, baseline, and bounded budget

Benchmark Optimization Loop by the numbers

  • 1,304 all-time installs (skills.sh)
  • +87 installs in the week ending Aug 5, 2026 (Skillselion tracking)
  • Ranked #886 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/affaan-m/ecc --skill benchmark-optimization-loop

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs1.3k
repo stars238k
Last updatedAugust 5, 2026
Repositoryaffaan-m/ecc

How do you optimize code performance with measured benchmarks?

Systematically test and improve code performance through measured, bounded optimization loops instead of vague speed requests.

Who is it for?

Backend engineers optimizing latency, throughput, memory, or cost who need bounded benchmark loops instead of unmeasured rewrite attempts.

Skip if: Greenfield feature development with no performance baseline, UI-only polish work, or tasks where correctness testing infrastructure does not yet exist.

When should I use this skill?

A developer asks to make something faster, benchmark latency or throughput, try optimization variants, or choose the best implementation by repeated measured tests.

What you get

Benchmark baseline report, variant comparison results, and selected best implementation with measured metric improvements.

  • Benchmark baseline
  • Variant comparison table
  • Selected optimized implementation

Files

SKILL.mdMarkdownGitHub ↗

Benchmark Optimization Loop

Use this skill to convert "make it 20x faster" or "try 50 recursive optimizations" into a bounded measured loop that can actually improve a system.

Required Baseline

Do not optimize until these exist:

  • the operation being optimized;
  • the correctness gate that must stay green;
  • the metric: wall time, p95 latency, rows/sec, cost/run, memory, error rate;
  • the current baseline;
  • the search budget: max variants, max time, max spend, max data impact.

If the user asks for an unrealistic target, keep the ambition but make the loop bounded and measurable.

Loop

1. Measure the baseline. 2. Identify bottlenecks from evidence. 3. Generate variants that test one hypothesis each. 4. Run variants with the same input shape. 5. Reject variants that fail correctness, safety, or reproducibility. 6. Promote the fastest safe variant. 7. Codify the winning path in a script, command, test, config, or doc. 8. Rerun the baseline and winner to confirm the delta.

Variant Table

Track variants like this:

Variant | Hypothesis | Command | Time | Correct? | Notes
baseline | current path | npm run job | 120s | yes | stable
batch-500 | fewer round trips | npm run job -- --batch 500 | 42s | yes | winner
parallel-8 | more workers | npm run job -- --workers 8 | 31s | no | rate limited

Recursive Search

For recursive or hyperparameter work:

  • persist every run to a ledger;
  • compare against the prior accepted winner, not only the previous run;
  • keep a holdout or replay check;
  • stop when improvement is within noise, correctness fails, cost exceeds the

budget, or the search starts changing more variables than it can explain.

Use phrases like "best measured safe variant" instead of "global optimum" unless the search space was actually exhaustive.

Promotion Gate

A variant cannot become the new default until:

  • correctness tests pass;
  • the performance delta is repeated or explained;
  • rollback is obvious;
  • the change is encoded in source control or a durable runbook;
  • the final summary includes exact commands and measurements.

Related skills

FAQ

What baseline is required before benchmark-optimization-loop runs?

benchmark-optimization-loop requires three prerequisites: the operation being optimized, a correctness gate that must stay green, and a chosen metric such as wall time, p95 latency, rows/sec, cost/run, or memory usage.

Which metrics does benchmark-optimization-loop track?

benchmark-optimization-loop supports wall time, p95 latency, rows per second, cost per run, and memory as optimization targets. Each variant is measured against the baseline while the correctness gate remains passing.

AI & Agent Buildingagentsautomation

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.