メニュー
Build ONNX Runtime from source. Use this skill when asked to build, compile, or generate CMake files for ONNX Runtime.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/microsoft/onnxruntime --skill ort-buildコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Trigger, re-run, and unblock ONNX Runtime CI checks on a GitHub pull request. Use this skill when a required check is stuck, missing, failed, or needs re-running. Nearly all ORT CI runs as GitHub Actions workflows; only the "Linux Android Emulator QNN CI Pipeline" remains on Azure Pipelines, plus the bot-driven "license/cla" and "Python format" lint checks. Triggers on: rerun CI, retrigger checks, stuck check, missing pipeline, failed CI, license/cla, Python format failure, Doc Gen CI failure, operator docs out of date, /azp run.
Draft ONNX Runtime release notes using preset configurations for full ORT or scoped component releases. Use when generating highlights with PR links, compiling human contributor acknowledgments from compile_contributors.py output, and applying preset path filtering.
Review ONNX Runtime pull requests, branches, commits, patches, and working-tree changes for actionable findings and missing tests.
SKILL.md を表示中
SOC 職業分類に基づく
| name | ort-build |
| description | Build ONNX Runtime from source. Use this skill when asked to build, compile, or generate CMake files for ONNX Runtime. |
The build scripts build.sh (Linux/macOS) and build.bat (Windows) delegate to tools/ci_build/build.py.
Three phases, controlled by flags:
--update — generate CMake build files--build — compile (add --parallel to speed this up)--test — run testsFor native builds, if none are specified (and --skip_tests is not passed), all three run by default. For cross-compiled builds, the default is --update + --build only.
--updateYou need --update when:
You do not need --update when only modifying existing .cc/.h files — just use --build. Skipping it saves time.
# Full build (update + build + test)
./build.sh --config Release --parallel
.\build.bat --config Release --parallel # Windows
# Just regenerate CMake files
./build.sh --config Release --update
./build.sh --config Release --build --parallel
./build.sh --config Release --
./build.sh --config Release --parallel --use_cuda --cuda_home /usr/local/cuda --cudnn_home /usr/local/cuda
./build.sh --config Release --parallel --build_wheel
./build.sh --config Release --build --parallel --target onnxruntime_common
./build.sh --build --parallel
| Flag | Description |
|---|---|
--config | Debug, MinSizeRel, Release, or RelWithDebInfo |
--parallel | Enable parallel compilation (recommended) |
--skip_tests | Skip running tests after build |
--build_wheel | Build the Python wheel package |
--use_cuda | Enable CUDA EP. Requires --cuda_home/--cudnn_home or CUDA_HOME/CUDNN_HOME env vars. On Windows, only cuda_home/CUDA_HOME is validated. |
--target T | Build a specific CMake target (requires --build; e.g., onnxruntime_common, onnxruntime_test_all) |
--use_webgpu | Enable WebGPU EP. To run its tests locally on Linux without a GPU, see the webgpu-local-testing skill. |
--cmake_extra_defines onnxruntime_QUICK_BUILD=ON | Faster CUDA build: instantiates a reduced kernel set. Side effect: Flash is compiled for head_dim 128 only, so most attention shapes fall back to MEA (changes which attention kernel is compiled/dispatched). Don't use it to characterize Flash-vs-arch behavior. |
--build_dir | Build output directory |
Default: build/<Platform>/<Config>/ where Platform is Linux, MacOS, or Windows.
With Visual Studio multi-config generators, the config name appears twice (e.g., build/Windows/Release/Release/).
It may be customized with --build_dir.
AGENTS.md.ort-test skill → "Verify which path/kernel actually executed"). Concrete instance:
onnxruntime_QUICK_BUILD=ON compiles FlashAttention for head_dim 128 only, so most
attention shapes silently dispatch to Memory-Efficient Attention instead of Flash —
details in the cuda-attention-kernel-patterns skill.python tools/ci_build/build.py directly over build.bat/build.sh when redirecting output. The .bat wrapper runs in cmd.exe, which breaks PowerShell redirection.> build_log.txt 2>&1). Build output is large and will overflow terminal buffers."Build complete" or errors.--parallel by default unless the user says otherwise.