403 个 · 按相关度排序
Profile and optimize Shiny application performance using profvis, bindCache, memoise, async/promises, debounce/throttle, and ExtendedTask for long-running computations. Use when the app feels slow or unresponsive during user interaction, when server resources are exhausted under concurrent load, when specific operations create bottlenecks, or when preparing an app for production deployment with many concurrent users.
中文简介正在准备中。当前显示原文说明。
Add Rcpp or RcppArmadillo integration to an R package for high-performance C++ code. Covers setup, writing C++ functions, RcppExports generation, testing compiled code, and debugging. Use when an R function is too slow and profiling confirms a bottleneck, when you need to interface with existing C/C++ libraries, or when implementing algorithms (loops, recursion, linear algebra) that benefit from compiled code.
中文简介正在准备中。当前显示原文说明。
Use when debugging CUDA with cuda-gdb or Compute Sanitizer, reading GPU core dumps, using device printf, or triaging error codes 700, 701, 702, and 719. Not for performance: use cuda-profiling.
中文简介正在准备中。当前显示原文说明。
Use when profiling with Intel VTune or AMD uProf for hotspots, top-down pipeline stalls, memory-bound analysis, or roofline data. Not for raw perf stat counters: use linux-perf.
中文简介正在准备中。当前显示原文说明。
Use when analyzing warp divergence, memory coalescing, shared memory bank conflicts, cache behavior, atomics, or occupancy tradeoffs on NVIDIA and AMD GPUs. Not for tool commands: use cuda-profiling.
中文简介正在准备中。当前显示原文说明。
Use when writing HIP kernels with hipcc, porting CUDA code through HIPIFY, profiling with rocprofv3, debugging with rocgdb, or optimizing for MI300X. Not for NVIDIA-side kernels: use cuda.
中文简介正在准备中。当前显示原文说明。
Use when tracing kernel functions with ftrace or trace-cmd, profiling with perf, kprobes, or dyndbg, or analyzing a vmcore with crash. Not for QEMU GDB stubs: use qemu-for-kernel-development.
中文简介正在准备中。当前显示原文说明。
Use when tuning or profiling native code on Apple M-series Macs: unified memory, 16 KiB pages, Accelerate and Metal for matrix work, xctrace and leaks, Rosetta 2, or sysctl hardware queries.
中文简介正在准备中。当前显示原文说明。
Use when adding telemetry, composing an observability surface, reviewing alerts, shipping a production feature, or diagnosing a production issue. Not for live failures, profiling, or launch runbooks.
中文简介正在准备中。当前显示原文说明。
Use when a rendering path needs stable frame-time, CPU, GPU, and memory evidence against fixed targets. Not for one-shot profiling or visual quality review.
中文简介正在准备中。当前显示原文说明。
Use when asked to build or review a CLI intended for coding agents and return flag-driven, pipeline-safe, idempotent design advice. Not for running or profiling a CLI: use control-cli.
中文简介正在准备中。当前显示原文说明。
Use when asked to optimize code, make a path faster, make this as fast as possible, reduce allocations, repair a regression, profile a target, grill every inefficiency, or run a subsystem or repo-wide performance campaign against a measured floor; when a performance requirement, slowness report, Core Web Vitals miss, or profiling evidence identifies a bottleneck; or for an "extremely optimize" performance campaign, optimizing suspected hot paths without waiting for benchmarks, or estimating hot and complexity-neutral cold paths while refusing complexity theater. Not for remote, credential, publish, deploy, or irreversible changes.
中文简介正在准备中。当前显示原文说明。
Diagnose React + Three.js (or any browser canvas) frame stutter by capturing a CDP CPU profile via Playwright instead of guessing from `[perf]` console logs. Use when: (1) frame time is bad but `[perf]` instrumentation doesn't pinpoint the offender, (2) you've spent more than ~30 minutes "fixing" suspects without measuring, (3) hot paths involve animation systems, allocation bursts, or invisible scene-graph traversal, (4) you need self-time per function with line numbers, not aggregate "render slow". The `console.warn` pattern is good for monitoring; CDP is needed for diagnosis. Covers script template, sourcemap setup so minified function names are decoded, and interactive variant where the user drives the interaction and signals stop via `touch /tmp/<flag>`.
中文简介正在准备中。当前显示原文说明。
专门用于 Ascend 集群 Profiling 性能数据的“快慢卡”诊断专家技能。当用户提供【集群性能数据目录/路径】并要求分析【快慢卡】、【慢节点】、【负载不均衡】或【集群瓶颈】时，必须触发此技能。该技能会自动接收集群路径，调度相关工具输出快慢卡的宏观定性与微观根因（如 Host 下发瓶颈、算子计算劣化）。
中文简介正在准备中。当前显示原文说明。
RL 强化学习训练与推理（rollout）路径的算子差异性扫描。当用户提供训练启动脚本路径，要求对比 Megatron 训练路径与 vLLM 推理路径的算子差异（尤其是融合算子 vs 单算子不一致）时使用此 skill。典型触发词包括"训推算子扫描"、"训推差异性"、"算子差异报告"、"融合算子对比"、"train vs infer op diff"。注意：本 skill 通过运行完整 RL 训练脚本（集成 profiling）来采集真实运行时算子（非静态源码扫描，非独立分离运行）。当训练、推理使用不同脚本时禁止触发。
中文简介正在准备中。当前显示原文说明。
Analyze Ascend NPU computation-side profiling data for single-card runs or a selected rank from multi-card runs. Use this skill when the user asks to diagnose computation bottlenecks, AI Core / AI Vector / AICPU hotspots, dynamic shape overhead, block dim issues, redundant TransData/Transpose/Cast, cross-stream waits, frequency/runtime-state issues, or fusion opportunities.
中文简介正在准备中。当前显示原文说明。
Interactively gather parameters for `python -m cli.inference.text_generate`, generate a concrete text generation simulation command, explain assumptions, ask for execution confirmation, then run the command and summarize performance metrics. Use when the user wants to validate a specific model, hardware profile, batch/query size, prefill or decode scenario, TP/DP/EP/MOE parallel strategy, profiling database, chrome trace, empirical metrics export, or a best row produced by throughput_optimizer.
中文简介正在准备中。当前显示原文说明。
MindStudio Profiler Analyze（msprof-analyze）是面向 AI 训练与推理场景的性能分析工具，基于采集得到的 profiling 数据进行统计、比对和诊断，帮助定位计算、通信、调度及集群场景下的性能瓶颈。
中文简介正在准备中。当前显示原文说明。
Analyze Ascend NPU collective communication profiling data with a DB-first workflow. Use when the user provides `cluster_analysis_output/cluster_analysis.db`, rank-level `analysis.db`, rank-level `ascend_pytorch_profiler_{rank_id}.db`, together with `profiler_info.json`, and asks about HCCL or hcom communication cost, collective communication TOP ops, wait time, slow rank/straggler, Notify Wait, bandwidth, retry, relay, SDMA/RDMA/HCCS links, communication matrix, or Ascend communication fault patterns.
中文简介正在准备中。当前显示原文说明。
Ascend cluster comparison tool. Invoke when user asks to compare two cluster datasets (DB or TEXT), or to generate cluster_analysis_output from raw profiling data via msprof-analyze and then compare.
中文简介正在准备中。当前显示原文说明。
运行程序采集 Profiling 数据，再用 msprof-analyze 分析TFLOPS 与 MFU。
中文简介正在准备中。当前显示原文说明。
Go software engineering skill for designing, implementing, debugging, testing, reviewing, and refactoring Go applications, CLIs, APIs, services, concurrency code, modules, and tooling. Use for .go files, go.mod/go.sum, goroutines/channels, interfaces, error handling, performance profiling, race fixes, and idiomatic Go architecture.
中文简介正在准备中。当前显示原文说明。
Iterative measurement-driven improvement loop. Measure, profile, mutate, re-measure, commit. Works for performance, bundle size, complexity, test coverage — anything quantifiable. Use when the user wants to systematically improve a metric through repeated cycles of profiling and targeted changes.
中文简介正在准备中。当前显示原文说明。
Optimize multi-agent systems with coordinated profiling, workload distribution, and cost-aware orchestration. Use when improving agent performance, throughput, or reliability.
中文简介正在准备中。当前显示原文说明。