一键导入
run-llvm-tests
Builds and runs LLVM/MLIR/Clang tests. Use when you need to verify changes to the LLVM project by running relevant test suites.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Builds and runs LLVM/MLIR/Clang tests. Use when you need to verify changes to the LLVM project by running relevant test suites.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Asks multiple AI models the same question via Cursor agents and synthesizes their answers. Use when you want diverse perspectives on a codebase question, architecture exploration, or code review.
Multi-model code review with two rounds — initial review, triage/fix, then rebuttal review. Orchestrates cursor agents to review your active changes from multiple perspectives. Use when the user says "let the peanut gallery review", "peanut gallery review", "cross-review", "multi-review", "multi-model review", or asks for a review from multiple agents/models.
| name | run-llvm-tests |
| description | Builds and runs LLVM/MLIR/Clang tests. Use when you need to verify changes to the LLVM project by running relevant test suites. |
Build all the required targets and run tests.
cd to itWORKSPACE_ROOT/buildninja check-mlir
ninja check-mlir-dialect-spirvninja check-llvm
LIT_FILTERenvironment variable with env, e.g.: env LIT_FILTER=AMDGPU ninja check-llvmMake sure that all test targets are actually built -- use ninja.
Summarize results