Local TensorZero documentation reference (latest). TensorZero documentation. Use for LLM gateway configuration, inference API, model provider integrations, optimization (fine-tuning, DICL), experimentation, evaluations, deployment, and observability.
The basics of how to program GPUs using Mojo. Use this skill in addition to mojo-syntax when writing Mojo code that targets GPUs or other accelerators. Use targeting code to NVIDIA, AMD, Apple silicon GPUs, or others. Use this skill to overcome misconceptions about how Mojo GPU code is written.
Refactor code to reduce complexity and duplication without changing behavior.
Rapidly onboard to pyfoma core internals (regex compiler + FST algorithms), make safe code changes, and avoid common semantic and performance regressions in fst.py, regexparse.py, atomic.py, algorithms.py, paradigm.py, and partition_refinement.py.
Optimize transformer attention with Flash Attention — 2-4x speedup, 10-20x memory reduction for long sequences on CUDA GPUs.
Add a new voice handler or feature to the voice module
Fict compiler and runtime engineering guide for correctness and reliability. Use when implementing or reviewing Fict compiler transforms, runtime reactivity behavior, SSR/resume flow, devtools/playground integration, or release CI gates in Fict repositories.
Generates Retry pattern for PHP 8.4. Creates resilience component with exponential backoff, jitter, and configurable retry strategies. Includes unit tests.
Generates Timeout pattern components for PHP 8.4. Creates execution time limit infrastructure with configurable timeouts, fallback support, stream timeouts, and unit tests.
Generates minimal, safe fixes for CI configuration issues. Provides fix templates for dependency, test, lint, infrastructure, Docker, and timeout failures.
CodeIgniter 4 framework knowledge base. Provides CI4 MVC architecture, DDD integration, persistence, services, security (Shield auth, Filters authorization, CSRF), event system, queue (codeigniter4/queue jobs, workers, retry), infrastructure components (cache, HTTP client, email, throttler), testing, and antipatterns for CodeIgniter PHP projects.
Generates Adapter pattern for PHP 8.4. Converts incompatible interfaces, wraps legacy code and external libraries. Includes unit tests.
Generates PSR-20 Clock implementation for PHP 8.4. Creates ClockInterface implementations including SystemClock, FrozenClock, and OffsetClock for time abstraction and testing. Includes unit tests.
Generates Flyweight pattern for PHP 8.4. Optimizes memory via shared intrinsic state. Includes unit tests.
Generates PSR-16 Simple Cache implementation for PHP 8.4. Creates CacheInterface with get/set/delete operations and TTL handling. Includes unit tests.
Initialize Claude Code optimization for a new project — detect stack, create memories, generate constitution, configure settings
Search and adapt Triton/HIP kernel patterns from a corpus to optimize AMD GPUs; use to find similar ops and reuse tiling/occupancy strategies.
This skill should be used when optimizing PyTorch models and kernels, including efficient tensor operations, torch.compile, custom autograd/CUDA/Triton extensions, mixed precision, memory and data pipeline tuning, model optimization techniques, CUDA graphs, and profiling.
CDNA3/MI300 HIP programming insights—chiplet/cache model, Infinity Cache, memory coherency, matrix cores, sparsity, and best practices.
This skill should be used when writing or tuning HIP kernels on AMD/NVIDIA GPUs, covering memory coalescing, shared-memory tiling, bank conflict avoidance, warp primitives, occupancy, vectorization, async ops, loop unrolling, and profiling.
This skill should be used when reasoning about GPU architecture fundamentals to guide kernel optimization choices such as memory hierarchy usage, execution model mapping, block sizing, and latency-aware tuning across HIP, Triton, and PyTorch.
This skill should be used when optimizing AMD GPU kernels on MI300 using the aiter project, including running op tests, benchmarking, iterating on kernel changes, and recording results in the kernel experiment database.
