tile-ai
Official@tile-ai
Enabling Lightning-Fast AI Workloads Development via Tiling
Agent Skills by tile-ai
Showing 21 vetted skills indexed across 2 GitHub repositories.
tile-aitilelang-pass-analyzer
Analyze TileLang Pass functionality and generate structured Markdown reports.
tile-aitilelang-op-design
Generate TileLang-Ascend operator design documents from requirements.
tile-aitilelang-pass-workflow-analyzer
Analyzes TileLang-Ascend Pass pipeline relationships and execution order for placement recommendations.
tile-aitilelang-op-generate
Generate TileLang-Ascend operator code and tests from design.md specifications.
tile-aitilelang-review-skill
Validate Python and C++ code formatting against CI rules and generate JSON/Markdown reports.
tile-aitilelang-mode-guide
Guide TileLang Ascend mode selection and pass_config configuration for NPU kernels.
tile-aitilelang-github-operations
Install GitHub CLI and create pull requests via authenticated workflows.
tile-aitilelang-debug-helper
Configures GDB and VSCode debugging for mixed Python-C++ TileLang Ascend examples.
tile-aitilelang-error-fixer
Diagnose and repair TileLang-Ascend Pass crashes via GDB backtraces, IR dumps, and AOT tests.
tile-aitilelang-submodule-pull
Pull tilelang submodules and third-party code with retries and logging.
tile-aitilelang-api-best-practices
Provide best practices for writing Ascend NPU kernels with TileLang API.
tile-ailifecycle-pull-request
Automate GitHub pull request lifecycle from draft creation to human review.
tile-aicreating-issue
Create validated GitHub issues with structured templates and gh CLI labels.
tile-ailifecycle-issue-fixer
Resolve GitHub issues via TDD, worktree isolation, and pull request creation.
tile-aitune-multiplication
Optimize GEMV, GEMM, and grouped GEMM kernels on NVIDIA Ampere and Hopper GPUs.
tile-aicommitting-changes
Enforce Git commit and branch workflows for the TileOPs repository.
tile-aitune
Profile TileOPs GPU kernel latency and bottlenecks with CUDA-event timings.
tile-aicheck-kernel-format
Validate kernel and op format conformance for T.prim_func and TileLang macros.
tile-aikernel-debug
Diagnose TileLang GPU kernel correctness failures across configurations and shapes.
tile-aimigrating-new-op
Migrate operators into TileOPs with a phased kernel-to-op workflow.
tile-aicreating-pull-request
Create draft GitHub pull requests with validated titles, bodies, and labels via gh CLI.
Frequently Asked Questions About tile-ai
FAQPage SchemaWhat specific tasks can engineers perform using these capabilities?▼
Engineers can design, generate, and debug Ascend NPU kernels, optimize GEMM and GEMV performance on NVIDIA hardware, and manage repository pull request lifecycles. The system provides structured validation for kernel formatting, latency profiling, and diagnostic support for complex operator crashes.
Which technical personas benefit most from these skills?▼
These skills are designed for kernel engineers, performance optimization specialists, and repository maintainers working on high-performance computing projects. It specifically supports developers focused on TileLang-Ascend integration and those maintaining large-scale GPU operator libraries.
What are the primary prerequisites for implementing these kernel optimizations?▼
Implementation requires an environment configured for TileLang-Ascend development, access to NVIDIA Ampere or Hopper hardware for profiling, and GDB for mixed-language debugging. Users must also maintain authenticated access to GitHub repositories to utilize the integrated lifecycle management features.