Coordinates verification of performance theories using profiling tools. Enforces Fermi verification protocol, runs appropriate tool skills, and synthesizes evidence from multiple sources. Use after generating theories to prove which issues actually matter. Use when validating hypotheses, running profiling tools systematically, or collecting quantitative evidence. (project)
Verifies performance theories using profiling tools and Fermi validation to prove which issues actually matter.
/plugin marketplace add antonykamp/cc-truffle-performance-plugin/plugin install antonykamp-cc-truffle-performance-plugin@antonykamp/cc-truffle-performance-pluginThis skill inherits all available tools. When active, it can use any tool Claude has access to.
WORKFLOW.mdSystematically verifies performance theories using profiling tools. Enforces rigorous methodology to ensure only proven issues make it into the final report.
Code analysis finds POTENTIAL issues. Tools PROVE which issues actually matter.
Input: List of theories from generating-performance-theories skill
Approach: Verify one theory at a time, highest-severity first. Fix it, re-profile, continue.
Output: Verified findings for generating-performance-reports
See WORKFLOW.md for detailed procedures.
| Purpose | Tool Skill |
|---|---|
| Hot function identification | profiling-with-cpu-sampler |
| Execution frequency | tracing-execution-counts |
| Optimization barriers | detecting-performance-warnings |
| Compilation behavior | tracing-compilation-events |
| Inlining analysis | tracing-inlining-decisions |
| Type stability | detecting-deoptimizations |
| Allocation patterns | profiling-memory-allocations |
| Deep IR analysis | analyzing-compiler-graphs |
Note on Compiler Graphs: When theories come from code analysis (e.g., "this allocation should be eliminated"), compiler graphs provide direct evidence of what the compiler actually did. Use them early for allocation/boxing theories, not as a last resort.
Predecessor: generating-performance-theories → Provides theories to verify
Successor: generating-performance-reports → Compiles verified findings
Tool Skills Used:
profiling-with-cpu-samplertracing-execution-countsdetecting-performance-warningstracing-compilation-eventstracing-inlining-decisionsdetecting-deoptimizationsprofiling-memory-allocationsanalyzing-compiler-graphs1. [establishing-benchmark-baseline] → Create baseline
2. [Run benchmarks] → Get timing data
3. [generating-performance-theories] → Generate theories
4. [verifying-performance-theories] → THIS SKILL
5. [generating-performance-reports] → Document findings
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
This skill should be used when the user asks to "create a hookify rule", "write a hook rule", "configure hookify", "add a hookify rule", or needs guidance on hookify rule syntax and patterns.