From 569899e8d9a1aacc8baae811b17b7e5cc9aa5851 Mon Sep 17 00:00:00 2001
From: Samit <285365963@qq.com>
Date: Tue, 28 Apr 2026 10:21:11 +0800
Subject: [PATCH] [Docs] Add performance profiling step to add-diffusion-model
 skill (#3196)

Signed-off-by: SamitHuang <285365963@qq.com>
---
 .claude/skills/add-diffusion-model/SKILL.md | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/.claude/skills/add-diffusion-model/SKILL.md b/.claude/skills/add-diffusion-model/SKILL.md
index 0b979e1a9..212e8b02c 100644
--- a/.claude/skills/add-diffusion-model/SKILL.md
+++ b/.claude/skills/add-diffusion-model/SKILL.md
@@ -544,6 +544,15 @@ Pipelines without `SupportsModuleOffload` fall back to scanning
 well-known attribute names (`transformer`, `text_encoder`, `vae`,
 etc.), which fails for non-standard names.
 
+### Step 12: Performance Profiling
+
+After verifying correctness and implementing parallelism/caching, profile the model's performance to identify bottlenecks and ensure optimal execution.
+
+See the [Profiling Single-Stage Diffusion](../../../docs/contributing/profiling.md#3-profiling-single-stage-diffusion) guide for detailed instructions on:
+1. Using the PyTorch profiler (`profiler: "torch"`) to capture detailed CPU/CUDA traces.
+2. Using Nsight Systems (`nsys`) with `profiler: "cuda"` for low-overhead CUDA traces.
+3. Controlling profiling via `omni.start_profile()` and `omni.stop_profile()`.
+
 ---
 
 ## Iterative Development Tips