Analyze a performance bottleneck prompt

Analyzes slow or resource-heavy code, forms hypotheses and proposes measurable improvements.

Ready prompt

You are an engineer experienced in performance analysis. Investigate the code below in light of the observed symptoms.

Runtime: 
Symptoms:


Code:
```

```

Output format:
1) Quick summary: the 3 most likely bottleneck hypotheses with short justifications.
2) Verification method per hypothesis: which profiler, metric and threshold.
3) Proposed improvements, from lowest risk to highest.
4) Expected impact (approximate) and possible side effects.
5) Modified code (single block) — only for the highest-impact change.
6) Regression test suggestion: how to measure the same load in the future.

Aim at measurable impact, not cosmetic change. State your assumptions explicitly.

Prompts are for illustration only. Accuracy isn't guaranteed—please read and adapt them for your situation.

“Open in …” sends the prompt to the provider. Don’t add sensitive data; their terms apply.

Usage tips

  • 1

    Attach profiler output (flamegraph, py-spy, perf) to move from guess to measurement.

  • 2

    State hardware limits (cores, RAM, disk type).

  • 3

    Mention single-user vs. high-concurrency — solutions differ.

This prompt is for general purposes. For legal, medical or financial decisions please consult a qualified professional.

Related prompts