Loading...
Loading...
Found 202 Skills
Diagnoses and optimises slow SQL queries using EXPLAIN ANALYZE. Covers identifying bottlenecks (sequential scans, bad estimates, heap fetches), index strategy, query rewrites, and verification. Invoked when the user asks to optimize a query, fix a slow database query, or improve database performance.
Iteratively optimize cuTile kernel performance through systematic profiling, bottleneck analysis, IR comparison, and targeted tuning. Covers tile sizes, occupancy, autotune configs, TMA, latency hints, persistent scheduling, num_ctas, flush_to_zero, and IR-level debugging. Use when asked to "optimize cutile kernel", "improve kernel perf", "tune cutile performance", "make kernel faster", or iteratively benchmark and refine a cuTile GPU kernel in the TileGym project.
Analyze host/CPU overhead in TensorRT-LLM inference from nsys traces. Detect whether host overhead is the bottleneck using GPU idle ratio, host prep exposed ratio, and per-phase evidence. For regressions, isolate forward steps via allreduce/NVTX patterns, compare host operation breakdowns across versions, and identify scheduling or request-management overhead. Supports optional inter-kernel gap, eager-vs-graph, pattern mapping, and multi-rank straggler drill-down. Use standalone or within perf-analysis. Triggers: host overhead, inter-step gap, scheduling overhead, forward step isolation, nsys iteration analysis, NVTX breakdown, request management overhead, GPU idle, host bottleneck, host prep exposed, inter-kernel gap, bubble analysis, graph coverage, eager kernel, rank imbalance, straggler detection.
Run an autonomous Humanize-governed SGLang SOTA performance loop for one LLM model: first perform the fixed fair SGLang/vLLM/TensorRT-LLM deployment search and benchmark, then start one RLCR loop that repeatedly decides the gap, profiles the current bottleneck, runs layer/kernel pipeline analysis, patches SGLang code, optionally uses ncu-report-skill for kernel evidence, and revalidates until SGLang matches or beats the best observed framework under the same workload and SLA.
Designs, reviews, and debugs DynamoDB data layers from design axioms — enumerates access patterns, chooses partition/sort keys and GSIs, decides single-table vs. multi-table, configures Streams, Global Tables, TTL, and zero-ETL integrations to OpenSearch/Redshift/SageMaker Lakehouse, and produces a defensible data-layer design with a monthly cost estimate and optional live validation. Applies whenever a user is designing, reviewing, or refactoring anything backed by DynamoDB — schemas, access patterns, GSIs, single- vs. multi-table choices, Streams consumers, transactional outboxes, Global Tables, zero-ETL pipelines — even when they don't say "axioms" or "design review." Also applies when debugging hot partitions, throttling, unbounded Scans, LWW conflicts, or surprise bills on DynamoDB workloads.
Find and fix game performance problems methodically — measure with the engine profiler first, reason about the frame-time budget, locate the CPU-vs-GPU bottleneck, then apply the right fix: object pooling, draw-call batching, fewer allocations/GC spikes, and asset budgets. Engine- neutral method that pairs with each engine's profiler. Use when the user mentions performance, optimize, low/dropping FPS, frame drops, stutter, lag, profiler, frame budget, draw calls, batching, garbage collection/GC spikes, object pooling, or "the game runs slow".
Use this skill when the user wants to analyze an existing pipe for improvement opportunities — automation gaps, manual bottlenecks, missing AI agents, field conditions, or adjacent processes. Acts as a process analyst: investigates, diagnoses, and improves the pipe in progressive rounds — each round delivers visible results.
Implement Mistral AI rate limiting, backoff, and request management. Use when handling rate limit errors, implementing retry logic, or optimizing API request throughput for Mistral AI. Trigger with phrases like "mistral rate limit", "mistral throttling", "mistral 429", "mistral retry", "mistral backoff".
A framework for classifying product decisions based on impact and reversibility. Use this when you feel like a bottleneck for your team, when you have a massive backlog of choices to make, or when you need to justify spending weeks of research on a single high-stakes problem.
Analyze VictoriaMetrics query trace JSON to diagnose slow queries and produce a structured performance report with time breakdown, bottleneck analysis, and optimization recommendations. ALWAYS use this skill when: (1) the user mentions a VictoriaMetrics or VM trace, query trace, or trace JSON, (2) the user provides or references a JSON file containing duration_msec/message/children fields, (3) the user asks why a VictoriaMetrics/VM query is slow and has trace output, (4) the user asks about vmstorage node distribution, cache misses, or rollup performance in the context of a trace, (5) the user mentions vmselect trace, trace=1, or query performance debugging with VictoriaMetrics. This skill provides a structured report template that ensures consistent, thorough analysis — do not attempt to analyze VM traces without it.
Specializes in analyzing Lynx trace data to diagnose performance issues and provide actionable optimization strategies. Key Scenarios: - Loading Performance: Diagnosing slow startup metrics (FCP, FMP, TTI) and white screen issues. - Smoothness Analysis: Investigating root causes for scroll jank, frame drops, and interaction lag. - Regression Detection: Comparing traces to identify performance degradation or verify optimization gains between versions. - Pipeline Deep Dive: Pinpointing bottlenecks in specific rendering stages like Layout, Paint, JS execution, and background threads. - Native Module Analysis: Investigating performance issues related to native module calls.
Author step content for Novu workflows defined in the Dashboard or generated/edited via the Novu MCP. Use when filling in step controls (subject, body, editorType, headers, body, conditions) for email, in-app, sms, push, chat, delay, digest, throttle, or HTTP Request steps.