deep-performance-investigation

Profile a hotspot with CPU sampler, warnings, deopt tracing, and IGV dumps.

Updated Aug 27, 2026

Search Tags:#performance #analysis #profiling #benchmark #graalvm #truffle #deoptimization

Ready to run

npx skills add https://github.com/antonykamp/cc-truffle-performance-plugin --skill deep-performance-investigation

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: deep-performance-investigation
Source: https://github.com/antonykamp/cc-truffle-performance-plugin/tree/main/skills/deep-performance-investigation
Command: npx skills add https://github.com/antonykamp/cc-truffle-performance-plugin --skill deep-performance-investigation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Targeted profiling to quantify the impact of uncertain performance findings in Truffle/GraalVM environments, enabling data-driven decisions rather than guesswork.

Core Features & Use Cases

Focused profiling of a single hotspot identified during broad-performance-investigation
Quantitative impact estimation with confidence levels to guide optimization priorities
Clear falsification path when data disproves a finding

Quick Start

Provide a specific low-confidence finding from broad-performance-investigation and run targeted profiling to estimate its impact.

Frequently Asked Questions about deep-performance-investigation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I estimate the impact of uncertain performance findings in GraalVM?▼

Targeted profiling estimates the impact of uncertain performance findings in GraalVM by applying focused CPU samplers, performance warnings, deopt tracing, and IGV dumps with precise filters to gather evidence and quantify hotspot impact.

How do I profile a specific Truffle hotspot to quantify its performance impact?▼

Profiling a specific Truffle hotspot involves applying focused tools like CPU samplers and deopt tracing with precise filters to gather evidence, returning an impact estimate with confidence or a clear falsification based on results.

What is deoptimization tracing and when do I need it for performance analysis?▼

Deoptimization tracing is a focused profiling tool used during performance analysis to track when GraalVM reverts optimized code, needed to gather evidence on uncertain findings and estimate their impact with confidence.

Can I falsify a low-confidence performance finding during targeted profiling?▼

Yes, targeted profiling can falsify a low-confidence performance finding by applying precise filters to tools like CPU samplers and IGV dumps, providing a clear falsification if the data disproves the finding.

What is the best way to analyze deoptimization issues in a Truffle environment?▼

The best way to analyze deoptimization issues in a Truffle environment is applying deopt tracing and IGV dumps with precise filters to gather evidence, returning an impact estimate with confidence or a clear falsification.