skills.rest
robotijn
robotijnCOMMUNITY
Explore all skills by this author →
GitHub

performance-profiler

Identify performance bottlenecks across CPU, memory, I/O, and query paths.

3|2|Updated Jan 23, 2026
Search Tags:#performance optimization#performance profiling#bottleneck analysis#continuous profiling#flame graphs#differential profiling#production observability
One-click install
npx skills add https://github.com/robotijn/ctoc --skill performance-profiler-robotijn
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: performance-profiler
Source: https://github.com/robotijn/ctoc/tree/main/skills/specialized/performance-profiler
Command: npx skills add https://github.com/robotijn/ctoc --skill performance-profiler-robotijn

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the guesswork of performance troubleshooting by identifying the exact root cause of slow endpoints, high CPU usage, memory leaks, lock contention, and inefficient database queries that cause user-facing latency and wasted infrastructure costs.

Core Features & Use Cases

  • Flame graph and differential profiling: Generates readable flame graphs for CPU, allocation, I/O, lock contention, and query performance issues, with differential views to attribute regressions directly to specific commits or deploys.
  • Multi-language production tooling: Supports pre-configured, low-overhead profiling pipelines for Python, Java, C#, C/C++, TypeScript/Node.js, and SQL, compatible with industry-standard tools like Pyroscope, async-profiler, and perf.
  • Use Case: If your checkout service's p99 latency spiked 200% after a recent deploy, use this Skill to pull a differential flame graph, identify the exact function responsible for the regression, and get a prioritized fix plan with expected impact estimates.

Quick Start

Use the performance-profiler skill to investigate the slow POST /api/checkout endpoint that has a 320ms p50 latency after the latest deploy, and generate a flame graph with differential attribution to the responsible commit.

Frequently Asked Questions about performance-profiler

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

FAQPage Schema
How do I find the root cause of high CPU usage or slow endpoint latency after a recent deploy?▼

Yes, it supports production observability and continuous profiling workflows for Python, Java, C#, C/C++, and TypeScript/Node.js stacks. It applies pre-configured, low-overhead profiling pipelines compatible with industry-standard tools like Pyroscope, async-profiler, and perf.

What is differential profiling and how does it attribute regressions to specific commits?▼

Differential profiling compares flame graphs across versions to isolate performance regressions. By generating differential views of CPU, allocation, and lock contention metrics, it attributes latency spikes directly to specific commits or deploys, eliminating guesswork in performance troubleshooting.

Can I analyze memory leaks and lock contention alongside database query performance?▼

Yes, you can analyze memory leaks, lock contention, and inefficient database queries simultaneously. The profiling process identifies and attributes bottlenecks across CPU, memory, I/O, and query execution paths, generating readable flame graphs and classifying findings using a standardized taxonomy.

Does this approach work for pre-deploy regression checks in backend systems?▼

Yes, it works for pre-deploy regression checks in backend systems and database workloads. The profiling mechanism supports ad-hoc incident response and continuous workflows, classifying findings by standardized taxonomy and producing actionable fix proposals with expected impact estimates.

How do I get actionable fix proposals with expected impact estimates for performance bottlenecks?▼

To get actionable fix proposals with expected impact estimates, run the profiling analysis across your CPU, memory, I/O, and query execution paths. The output classifies identified bottlenecks using a standardized taxonomy and generates a prioritized fix plan aligned with 2026 observability best practices.

Related Skills

Explore More Software Engineering Skills→
jabrenajabrena
423

161-java-profiling-detect

Profile Java applications with async-profiler v4.x to detect CPU, memory, lock, GC, and I/O bottlenecks.

Community
Advanced
bennybennisonbennybennison

eval-driven-development

Define expected outcomes and run pass@k evals to validate agent behavior changes.

Community
Intermediate
mr-daedaliummr-daedalium
107

land-and-deploy

Merge PRs, trigger deployment, and verify canary health in CI/CD pipelines.

Community
Advanced
jabrenajabrena
423

704-technologies-sql

Review SQL schema, migrations, queries, indexes, and database security practices.

Community
Advanced