性能
性能分析与优化
mind-clone
yzfly
Activate a Cognitive Digital Twin. Can simulate a custom subject (via core/ directory) OR load pre-installed celebrity/expert personas (from personas/ directory).
langchain-use
NanmiCoder
LangChain 1.0 使用指南。提供 Agent、Tool、Memory、Middleware 等核心概念的快速参考。当用户需要创建 AI Agent、集成 LangChain、或解决 LangChain 相关问题时激活。
Reducing Complexity
obra
Managing complexity is software's primary technical imperative - all other goals are secondary
anki
doasfrancisco
Create and browse Anki flashcards during coding sessions. Use when the user invokes /anki, asks to create a flashcard, or when you notice the user learning something new and worth remembering (a concept, trick, gotcha, or mental model).
memoryguard
irisxc4
Local-first MCP memory backend and governance console for coding agents. Auto-organize, quarantine, supersede, and rollback shared memories across multiple agents.
quantizing-models-bitsandbytes
ovachiever
Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limited, need to fit larger models, or want faster inference. Supports INT8, NF4, FP4 formats, QLoRA training, and 8-bit optimizers. Works with HuggingFace Transformers.
web-browser
sourcegraph
'Allows to interact with web pages by performing actions such as clicking buttons, filling out forms, and navigating links. It works by remote controlling Google Chrome or Chromium browsers using the Chrome DevTools Protocol (CDP). When Claude needs to browse the web, it can use this skill to do so.'
remotion-performance-optimizer
ncklrs
Analyzes Remotion compositions for performance issues and provides optimization recommendations. Identifies expensive computations, unnecessary re-renders, large assets, memoization opportunities, and architecture improvements. Use when optimizing render times or when asked to "improve performance", "speed up renders", "optimize Remotion video".
performance-marketer
ncklrs
Expert performance marketing guidance for paid acquisition, CRO, and growth optimization. Use when creating ad campaigns, writing ad copy, optimizing landing pages, setting up A/B tests, analyzing CAC/LTV, building attribution models, or planning retargeting strategies. Use for Google Ads, Meta Ads, LinkedIn Ads, conversion tracking, budget allocation, and funnel optimization.
product-manager
ncklrs
Expert product management guidance for day-to-day PM work. Use when creating roadmaps, prioritizing features, managing stakeholders, planning sprints, grooming backlogs, scoping features, planning releases, defining OKRs, managing technical debt, or coordinating go-to-market. Covers RICE, ICE, MoSCoW frameworks, cross-functional collaboration, and product metrics.
remotion-video-reviewer
ncklrs
Structured review process for Remotion video implementations. Analyzes spec compliance, detects common timing/easing issues, validates asset quality, and provides prioritized revision lists. Use when reviewing Remotion code against design specs or performing quality assurance on video compositions. Trigger phrases "review video code", "check spec compliance", "audit Remotion implementation".
sales-enablement
ncklrs
Expert sales enablement strategist for building high-performing sales teams. Use when designing sales training programs, onboarding and ramp plans, sales playbooks, coaching frameworks, certification programs, or competitive intelligence distribution. Covers content strategy, tool adoption, performance measurement, and continuous learning systems. Use for building sales academies, creating enablement content, and optimizing sales productivity.
local-ai-model-matcher
Maciejdziuba
Safely auto-detect the current machine's hardware (GPU, VRAM, system RAM, OS) with read-only commands, then recommend the top 3 local AI models to run — most powerful that fits, best for coding, and best small-and-fast — each with download links. Use when someone asks "what AI models can my computer run" or "what's the best local model for my rig".
stock-watcher
agentbay-ai
Manage and monitor a personal stock watchlist with support for adding, removing, listing stocks, and summarizing their recent performance using data from 10jqka.com.cn. Use when the user wants to track specific stocks, get performance summaries, or manage their watchlist.
role-algorithms:data-structures
rnavarych
Implements and selects optimal data structures — hash tables (chaining, open addressing, Robin Hood, cuckoo), balanced BSTs (AVL, Red-Black, B-trees), heaps (binary, Fibonacci, pairing), tries, skip lists, segment trees, Fenwick trees, Bloom filters, Count-Min Sketch, HyperLogLog, and Union-Find. Use when choosing structures for performance constraints, implementing custom collections, or optimizing memory access patterns.
xbird
checkra1neth
"Use when the user asks to tweet, post threads, read tweets, search Twitter/X, check mentions, manage engagement (like/retweet/bookmark), update profile (bio, avatar, banner), upload media, or interact with Twitter accounts. Triggers: twitter, tweet, post, thread, timeline, mentions, followers, following, likes, retweet, bookmark, profile picture, bio."
role-aqa:performance-testing
rnavarych
Performance test design and execution with k6 (JavaScript, cloud execution, thresholds), JMeter (GUI, distributed), Gatling (Scala DSL), and Artillery (YAML config). Load, stress, soak, and spike testing. Baselines, bottleneck analysis, CI performance gates, custom metrics, and result analysis.
dev-run-tests
SyFI-ServingStudio
Use when asked to run, select, or understand ServingStudio Sim's tests — which tier needs a GPU / the built binary / an agent runtime / is just CPU, how the auto-skip works, and how the per-GPU throughput & sim-speed goldens are recorded. Covers the just test-* recipes, pytest markers, and the GPU-type-tagged golden store. NOT for writing a new kernel's cache-fidelity check (that is impl-validate-kernel-cache).
dev-llm-serving
SyFI-ServingStudio
Implement or review code in a real LLM or multimodal serving framework using the routed reference library. Under a parent workflow, follows its frozen trial.
operate-profile-sim-speed
SyFI-ServingStudio
Use when asked to profile, perf, or speed up the ServingStudio Sim simulator wallclock (how fast a run executes), not the modeled cluster throughput and not L1 kernel profiling. Covers building a symbol-rich release binary, capturing a perf record of a representative run, and reading the per-thread / flat breakdown to find the bottleneck.
vwap-volume-profile
johnamcruz
Use session and RTH VWAP, the volume profile (POC, value area, high and low volume nodes), and relative volume to judge trend quality, fair value, and participation in futures. Use when evaluating mean-reversion vs trend days, where price sits against value, targets at volume nodes, or whether a breakout has participation.
top-compose-real-framework-from-sim
SyFI-ServingStudio
Build or optimize a real serving framework through repeated simulation, one measured trial, and the Align framework to ServingStudio Sim side of operate-run-alignment.
top-split-model-into-kernels
SyFI-ServingStudio
Use as the entry point when the user wants to break a Transformer model's forward pass into the sequence of ServingStudio Sim kernels it costs as — the per-op boundary decision. First calls top-explore-models to establish the architecture, then for each op routes to dev-lookup-transformers-model (what math) and dev-explore-kernel (whether a real fused kernel exists at what launch granularity), then applies the boundary rules to assign each op a home: reuse an existing kernel, a new dedicated kernel (hand to top-add-kernel), an elementwise byte-placeholder, or fold into a neighbor. Never skips a cost-bearing op. When a measured vLLM/SGLang capture is available it is the primary evidence for launch granularity, and the table is ordered by each op's measured share of iteration time. Produces a per-op decision table, not code.
proxmox-admin
Bldg-7
Proxmox Virtual Environment operational expertise for AI agents - VM lifecycle, LXC containers, storage management, backup strategies, HA configuration, and troubleshooting