From 62b0da9e1be796a9c34b498f306d9e12e2fd781b Mon Sep 17 00:00:00 2001
From: Viet-Anh Nguyen <vietanh@nrl.ai>
Date: Sun, 30 Aug 2026 02:21:47 +0700
Subject: [PATCH] docs: require cross-platform artifact validation

---
 TODO.md | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/TODO.md b/TODO.md
index e39653f..44bfd42 100644
--- a/TODO.md
+++ b/TODO.md
@@ -1,5 +1,9 @@
 # TODO
 
+- [ ] Before each binary release, build from clean, dedicated AnyLabeling
+  environments and checksum/launch-test the exact uploaded CPU and GPU
+  artifacts on Linux, Windows, and Apple Silicon. Run real accelerator
+  inference on each available CUDA, DirectML, CoreML, OpenVINO, or NPU device.
 - [ ] Validate real inference in isolated accelerator environments on Linux
   NVIDIA CUDA, Windows CUDA and DirectML, and Apple Silicon CoreML after every
   accelerator-runtime change.