Claude Skills by curiositech

Integration of Kahneman's heuristics research with Klein's expertise studies on intuitive decision-making

IBM's autonomic computing vision for self-managing systems with self-configuration, healing, optimization, and protection

Declarative programming framework for optimizing LLM prompts through compilation and automatic tuning

GOMS cognitive modeling methodology for analyzing human-computer interaction task performance

Naturalistic decision-making research showing how experts make decisions through recognition-primed processes

**Source**: *Designing Data-Intensive Applications* by Martin Kleppmann **Domain**: Distributed systems, data architecture, reliability engineering **Applies to**: Building systems where data complexity (not computation) is the bottleneck

Expert in visualizing and explaining braid theory, knot mathematics, and topological concepts for educational purposes. Use for creating interactive visualizations, explainer cards, step-wise animations, and translating abstract algebra into intuitive understanding. Activate on keywords: braid theory, knot visualization, σ notation, crossing diagrams, Yang-Baxter, topological education. NOT for general math tutoring, pure knot invariant computation, or non-educational knot theory research.

Choose what kind of knowledge to transfer between teacher and student models: response, feature, or relational, and decide among offline, online, self, or cross-modal distillation schemes. Best for distillation strategy selection, capacity-gap diagnosis, and transfer planning. Activate on "knowledge distillation", "teacher-student", "soft labels", "feature distillation", "online distillation", or "cross-modal transfer". NOT for generic compression checklists or unrelated training work.

Deep analysis of knowledge distillation techniques for compressing large models into smaller efficient ones

Design deployment-focused distillation systems that balance model size, accuracy, calibration, and cascade escalation under real resource limits. Best for teacher-student compression, threshold design, and failure-aware deployment. Activate on "model compression", "teacher- student", "distillation score", "cascade model", "edge deployment", or "model calibration". NOT for generic deep-learning overviews, prompt optimization, or training work without a concrete distillation objective.

Comprehensive survey of knowledge distillation methods, architectures, and applications in neural network compression

license: Apache-2.0 NOT for unrelated tasks outside this domain.

Use when triaging CrashLoopBackOff, OOMKilled, ImagePullBackOff, Pending pods, networking failures, ingress 502/504, PVC stuck, HPA not scaling, init-container loops, sidecar startup ordering, or "kubectl describe says nothing useful". Triggers: pod restart counts climbing, "no nodes available to schedule", "back-off restarting failed container", DNS resolution failures inside the cluster, NetworkPolicy denials, livenessProbe killing healthy pods. NOT for cluster setup (kubeadm, kops), cloud-...

Automate Kubernetes deployment lifecycle: kubectl rollouts, GitOps sync (ArgoCD/Flux), namespace and secret management, rollout monitoring, and failed deployment troubleshooting. Use when deploying manifests to clusters, debugging stuck rollouts, setting up GitOps pipelines, or managing multi-environment promotion. NOT for: generating K8s manifests (use kubernetes-manifest-generator), blue-green/canary traffic shifting (use blue-green-deployment-orchestrator), cluster provisioning (use terraf...

Shutting down HTTP services in Kubernetes without dropping traffic — the SIGTERM/preStop dance, the EndpointSlice removal race, the `preStop sleep` pattern, terminationGracePeriodSeconds budgeting, and the Node.js + Go server skeletons. Grounded in kubernetes.io and language runtime docs.

Kubernetes manifest and Helm chart generator for production workloads. Activate on: K8s config, Deployment YAML, HPA autoscaling, PodDisruptionBudget, Ingress rules, NetworkPolicy, Helm chart, Kustomize. NOT for: Docker image building (use docker-multi-stage-optimizer), IaC provisioning of clusters (use terraform-module-builder), CI/CD pipeline config (use github-actions-pipeline-builder).

Philosophy of science methodology examining how research programs evolve through proofs and refutations

Delta Lake, Apache Iceberg, Hudi for ACID transactions on object storage. Activate on: lakehouse, Delta Lake, Iceberg, Hudi, table format, ACID on S3, time travel, data lake, open table format. NOT for: warehouse query tuning (use data-warehouse-optimizer), streaming ingestion (use streaming-pipeline-architect).

Master of high-performance web map implementations handling 5,000-100,000+ geographic data points. Specializes in Leaflet.js optimization, Supercluster algorithms, viewport-based loading, canvas rendering, and progressive disclosure UX patterns.

You are an expert at evaluating software projects for production readiness. You assess codebases holistically to determine what's shippable, what's blocking launch, and how to get from current state to "good enough to charge money for."

license: Apache-2.0 NOT for unrelated tasks outside this domain.

Most people's mental model of expert decision-making focuses on choice: the expert sees the options clearly and selects the best one. The NDM research tradition, and Zimmerman's empirical study, reveals that this model is inverted. The expert's primary cognitive investment is not in choosing — it is in understanding. The action follows almost automatically from a correct situational assessment.

Writes legal-tech advertorials, landing pages, and blog-to-CTA copy for expungement, debt relief, housing rights, and related services. Use when turning educational legal content into honest, trauma-informed conversion copy with clear competitive framing and non-hyped CTAs. NOT for legal advice, courtroom filings, or manipulative direct-response copy that overstates results.

Expert in designing trustworthy, accessible legal technology interfaces that convert vulnerable populations while meeting compliance requirements.

Human interface agent that translates ecosystem activity into clear, actionable communication. Creates status briefings, decision requests, celebration reports, concern alerts, and opportunity summaries. Use for 'status update', 'brief me', 'what's happening', 'summarize progress', or when complex multi-agent work needs human-readable reporting.

Comprehensive benchmark suite for evaluating LLM agents across diverse interactive environments

Using LLMs as automated evaluators for comparing chatbot responses with Chatbot Arena methodology

Track and reduce LLM API costs with per-request token tracking, model routing, budget alerts, and prompt compression. Activate on: reduce AI costs, token tracking, model routing, LLM budget, prompt compression. NOT for: general cloud cost optimization (cost-accrual-tracker), model training costs (ai-engineer).

Build automated LLM evaluation pipelines with benchmarks, regression tests, RAGAS, and human eval workflows. Activate on: LLM evaluation, benchmark testing, eval pipeline, RAGAS, model regression tests. NOT for: traditional software testing (testing-expert), model training (ai-engineer).

Implement semantic and exact-match caching for LLM responses to reduce cost 40-60% and latency. Activate on: LLM caching, semantic cache, reduce API costs, cache AI responses. NOT for: general web caching (caching-strategies), CDN config (cloudflare-worker-dev).

Selects the optimal LLM model and provider for each task based on complexity, cost budget, and capability requirements. Routes cheap tasks to Haiku/GPT-4o-mini and complex tasks to Sonnet/Opus/o1. Use when deciding which model to call, optimizing LLM costs, or building multi-model agent systems. Activate on "which model", "model selection", "route to model", "LLM cost", "model routing", "cheap vs expensive model". NOT for prompt engineering (use prompt-engineer), model fine-tuning, or trainin...

Build production LLM streaming UIs with Server-Sent Events, real-time token display, cancellation, error recovery. Handles OpenAI/Anthropic/Claude streaming APIs. Use for chatbots, AI assistants, real-time text generation. Activate on "LLM streaming", "SSE", "token stream", "chat UI", "real-time AI". NOT for batch processing, non-streaming APIs, or WebSocket bidirectional chat.

Centralized log pipeline architect with structured logging, Fluentd/Vector, and retention policies. Activate on: log aggregation, structured logging, Fluentd, Vector, Loki, ELK stack, log pipeline, log retention, centralized logging. NOT for: metrics and dashboards (use monitoring-stack-deployer), distributed tracing (use logging-observability), alerting rules (use site-reliability-engineer).

Structured logging, distributed tracing, and metrics for production applications. [What: OpenTelemetry setup, log level strategy, correlation IDs, SLI/SLO alerting thresholds, Grafana dashboard design, PagerDuty integration] [When: setting up production logging, adding observability to a service, debugging distributed systems, designing alerting, implementing traces/metrics/logs] [Keywords: logging, observability, OpenTelemetry, OTel, structured logs, distributed tracing, correlation ID, metr...

Identify, label, and explain logical fallacies in arguments and discourse. Use when the user wants to analyze arguments, check reasoning, evaluate debate claims, spot weak arguments, do debate prep, policy writing, or conflict analysis. Triggers on 'fallacy', 'logical error', 'faulty reasoning', 'argument flaw', 'is this valid', 'check this logic'. Works on both formal deductive errors and informal reasoning mistakes.

Plan and run manager-led, role-based teams where assignments change by round, roles may be added midstream, and the manager decides when the work is ready to ship. Use when delegation should adapt to emerging evidence rather than a fixed review loop. NOT for static approval chains, uncoordinated message swarms, or blackboard-style shared-state diagnosis.

license: Apache-2.0 NOT for unrelated tasks outside this domain.

Expert in maximalist interior wall decoration including bold color choices, freehand paintings, statement wallpapers, eclectic lamp arrangements, gallery walls, and curated chaos. Embraces "more is more" philosophy with sophisticated color theory and composition. Activate on "wall decor", "maximalist design", "bold colors", "gallery wall", "statement wallpaper", "freehand painting", "eclectic style", "accent wall", "lamp collection", "more is more", "silly decor". NOT for minimalist design (d...

Creates web experiences that embrace density, richness, and visual abundance. Maximalism is the deliberate, skilled rejection of "less is more" — it argues that **more is more**, that empty space is wasted space, and that visual complexity can be navigated when properly orchestrated.

Expert MCP (Model Context Protocol) server developer creating safe, performant, production-ready servers with proper security, error handling, and developer experience. Activate on 'create MCP', 'MCP server', 'build MCP', 'custom tool server', 'MCP development', 'Model Context Protocol'. NOT for using existing MCPs (just invoke them), general API development (use backend-architect), or skills/agents without external state (use skill-coach/agent-creator).

Use when designing, building, or debugging a Model Context Protocol (MCP) server in Node/TypeScript. Triggers: stdio JSON-RPC handshake, tool descriptions as discovery surface, lazy startup vs eager catalog loading, telemetry placement, schema design for tool inputs (zod), tool naming conventions for discoverability, error handling that does not leak stack traces, MCP client compatibility (Claude Desktop, Claude Code, Cursor), local resource fetching, secrets and env var handling, distributin...

Comprehensive MDX content sanitizer that escapes angle brackets, generics, and other JSX-conflicting patterns to prevent build failures

Design agent-labor markets with bond pricing, escrow settlement, oracle incentives, cold-start bootstraps, and collusion-resistant marketplace rules. Use for collateralized task contracts, Float Plan bond sizing, settlement design, and incentive analysis for agent work. NOT for general auction theory teaching, DeFi tokenomics, payment gateway integration without marketplace incentives, or generic game theory tutoring.

license: Apache-2.0 NOT for unrelated tasks outside this domain.

Renders Mermaid diagrams to SVG, PNG, and PDF in both web and offline contexts. Covers client-side lazy loading, SSR trade-offs, CLI batch export, and native Rust rendering. Use when exporting diagrams to images, optimizing Mermaid rendering performance on web pages, setting up CI/CD diagram pipelines, or choosing a rendering library. Activate on "render mermaid", "export diagram", "mermaid to png", "mermaid to svg", "mermaid performance", "diagram export". NOT for writing Mermaid syntax (use...

Writes precise, well-structured Mermaid diagrams for any visualization need. Use when creating flowcharts, sequence diagrams, state machines, ER models, timelines, mindmaps, Gantt charts, or any other Mermaid-supported diagram type. Activate on "mermaid", "diagram", "flowchart", "sequence diagram", "state diagram", "ER diagram", "visualize", "draw graph". NOT for rendering/exporting Mermaid to images (use mermaid-graph-renderer), ASCII art, or GUI-based design tools.

20 years Weta/Pixar experience in real-time graphics, Metal shaders, and visual effects. Expert in MSL shaders, PBR rendering, tile-based deferred rendering (TBDR), and GPU debugging. Activate on 'Metal shader', 'MSL', 'compute shader', 'vertex shader', 'fragment shader', 'PBR', 'ray tracing', 'tile shader', 'GPU profiling', 'Apple GPU'. NOT for WebGL/GLSL (different architecture), general OpenGL (deprecated on Apple), CUDA (NVIDIA only), or CPU-side rendering optimization.

Decide HOW LOW to go for a bare-metal 2D/text rendering pipeline on Apple GPUs, and pay the right costs. Covers the layered choice — pure objc2-metal + hand-written MSL (own command queue, glyph atlas, frame pacing) vs wgpu (Metal under the hood, cross-platform) vs Vello-on-wgpu (compute vector renderer you don't have to write) — plus CAMetalLayer/CADisplayLink frame pacing, ProMotion 120Hz, the glyph-atlas / signed-distance-field text problem, CPU-GPU sync, and honest cost accounting of "pur...

Microservice architecture patterns — service decomposition, inter-service communication, API gateway, saga pattern, event-driven architecture, service mesh, circuit breaker, CQRS, event sourcing. Activate on "microservices", "service decomposition", "saga pattern", "API gateway", "event-driven", "service mesh", "circuit breaker", "CQRS", "event sourcing", "bounded context", "strangler fig", "distributed transactions", "choreography vs orchestration". NOT for monolith design, serverless functi...

Coaches end-to-end ML system design interviews covering inference pipelines, recommendation systems, RAG, feature stores, and monitoring. Use for L6+ design rounds, ML architecture whiteboarding, system design practice, serving tradeoff analysis. Activate on "ML system design", "ML interview", "recommendation system design", "RAG architecture", "feature store design", "model serving". NOT for coding interviews, behavioral questions, ML theory quizzes, or paper implementations.