| name | compiler-runtime-analysis |
| description | Analyze compiler, runtime, and code generation systems including IR lowering, scheduling, memory layout, graph compilation, autotuning, and runtime overhead. Use when profiling a compiler stack, comparing generated code, debugging performance cliffs, or evaluating compiler research claims. |
Compiler Runtime Analysis
Use this skill when performance claims depend on what the compiler and runtime actually did.
Core Workflow
- State the source program, target hardware, runtime model, and optimization goal.
- Trace the path from high-level program to lowered IR, generated code, and runtime launch behavior.
- Use
references/measurement-checklist.md before claiming a codegen win.
- Separate compilation time, runtime overhead, and steady-state kernel performance.
- Compare generated artifacts, not just wall-clock summaries.
Execution Rules
- Keep correctness and speed as separate gates.
- Check shape specialization, caching, and autotuning warmup effects.
- Distinguish compiler wins from library or kernel wins.
- Always name the target backend and input regime.
Output Contract
Return:
- Pipeline map.
- Measurement plan.
- Likely bottlenecks.
- Artifact comparison points.
- Recommended interventions.