Spring Boot testing fundamentals including @SpringBootTest, slice tests, MockMvc, mocking with @MockBean, test configuration, and testing utilities. For integration tests with Testcontainers, see spring-boot-integration. USE WHEN: user mentions "spring boot test", "mockmvc", "@WebMvcTest", "@DataJpaTest", asks about "@SpringBootTest", "@MockBean", "spring test", "controller test", "repository test" DO NOT USE FOR: Unit tests - use `junit` with Mockito; Integration tests with real DB - use `spring-boot-integration`; REST API clients - use `rest-assured`; E2E tests - use Selenium

SQLCipher — transparent AES-256 encryption for SQLite databases. Drop-in SQLite replacement used by mobile wallets, password managers, and offline-first apps storing sensitive data. Covers key derivation (PBKDF2), key rotation, performance tuning, integration in Rust (rusqlite + bundled-sqlcipher feature), Android (sqlcipher-android), iOS (SQLCipher.framework or Swift package), KMP (SQLDelight + SQLCipher driver), license model (community BSD-3 vs commercial). USE WHEN: user mentions "SQLCipher", "encrypted SQLite", "PRAGMA key", "PRAGMA rekey", "sqlite-cipher", "rusqlite bundled-sqlcipher", "encrypted database for mobile wallet", "SQLDelight encryption" DO NOT USE FOR: Plain SQLite without encryption - use `databases/sqlite` (or relevant skill) DO NOT USE FOR: Server-side encrypted DB (use TDE) - use `databases/postgresql` etc. DO NOT USE FOR: Encryption primitives outside SQLite - use `security/libsodium`

SSL/TLS certificate management skill: Let's Encrypt, Certbot, certificate types, TLS version/cipher configuration, HSTS, OCSP stapling, renewal automation, and local development certificates. USE WHEN: - Obtaining and renewing Let's Encrypt certificates (standalone, webroot, DNS challenge) - Configuring TLS on Nginx or Apache (protocols, cipher suites, session parameters) - Setting up HSTS with preload, OCSP stapling, or DHParam generation - Automating certificate renewal via systemd timer and deploy hooks - Managing multi-domain (SAN) or wildcard certificates - Debugging TLS handshake failures, mixed content, or ACME challenge errors - Setting up mkcert for trusted local HTTPS development DO NOT USE FOR: - mTLS / client certificate authentication (use service-mesh or api-gateway skill) - Application-level JWT or session token management - SSH key management - Internal CA for Kubernetes clusters (use cert-manager)

Tailwind CSS utility framework. Covers utility classes, responsive design, and customization. Use for styling web applications. USE WHEN: user mentions "Tailwind", "utility CSS", "Tailwind classes", asks about "responsive design", "Tailwind config", "utility-first CSS" DO NOT USE FOR: CSS-in-JS (styled-components, emotion), CSS Modules only projects, traditional CSS/SCSS workflows, component libraries like Material-UI

Traefik v3 cloud-native reverse proxy. Covers providers, entrypoints, routers, middlewares, services, Docker labels, TLS/ACME, dashboard, and metrics. USE WHEN: user mentions "traefik", "traefik v3", "traefik docker", "traefik labels", "traefik middleware", "traefik dashboard", "traefik tls", "traefik acme", "traefik router", "traefik entrypoint", "traefik reverse proxy", "traefik cloudflare", "traefik let's encrypt", "traefik rate limit" DO NOT USE FOR: Caddy-based setups - use `caddy` skill, Nginx load balancing - use `load-balancer` skill, Kubernetes ingress with nginx-ingress - use `kubernetes` skill, Application-level TLS inside app code

Zustand state management for React. Covers stores, actions, and persistence. Use for simple global state management. USE WHEN: user mentions "zustand", "global state", "store", asks about "simple state management", "lightweight state", "create store", "persist state", "middleware", "devtools integration" DO NOT USE FOR: server data - use `tanstack-query` or `swr` instead; Vue apps - use `pinia`; complex async workflows - use `redux-toolkit`

Apache ActiveMQ message broker with JMS support. Covers queues, topics, message selectors, and Spring integration. Use for enterprise Java messaging and JMS-compliant applications. USE WHEN: user mentions "activemq", "jms", "artemis", "message selectors", "virtual topics", asks about "java messaging", "jms queues", "enterprise messaging", "spring jms" DO NOT USE FOR: event streaming - use `kafka` or `pulsar`; cloud-native - use `nats`; AWS-native - use `sqs`; Azure-native - use `azure-service-bus`; non-JMS preferred - use `rabbitmq` or `kafka`

age — modern file encryption format and tool by Filippo Valsorda. Replaces GPG for most use cases (encrypted backups, exports, secrets in CI). Covers age CLI, X25519 + Scrypt-based recipients, SSH key recipients, plugin system (YubiKey, Secure Enclave, age-keyring), Rust (`age` crate), Go (filippo.io/age), encrypted backup workflows for wallets. USE WHEN: user mentions "age", "age-encryption", "rage", "filippo.io/age", "ssh-rsa age", "age plugin", "age-yubikey", "age recipient", "age identity", ".age file", "age-keygen" DO NOT USE FOR: SQLite encryption - use `databases/sqlcipher` DO NOT USE FOR: General crypto primitives - use `security/libsodium` DO NOT USE FOR: GPG-specific workflows - use GPG-specific tooling DO NOT USE FOR: Real-time stream encryption - use `security/libsodium` (secretstream)

Real-time RAG ingestion. CDC (Debezium, Postgres logical replication), Kafka/ Pulsar topics for doc events, stream processing (Flink, Kafka Streams) to embedding service, exactly-once semantics, late-arriving updates, tombstones (deletes), upsert to vector DB, schema evolution. Full Debezium + Kafka -> vector DB example. USE WHEN: user mentions "CDC RAG", "Debezium RAG", "Kafka RAG", "real-time embeddings", "streaming ingestion", "Flink embeddings", "Pulsar RAG", "logical replication RAG" DO NOT USE FOR: batch scheduled ingestion - use `ingestion-orchestration`; query-time freshness weighting - use `time-aware-retrieval`; evaluation - use `rag-evaluation`

Class-validator decorator-based validation for TypeScript. Standard in NestJS. Use when building APIs with NestJS or needing class-based validation. USE WHEN: user mentions "class-validator", "NestJS validation", "DTO validation", "decorator validation", asks about "NestJS DTOs", "ValidationPipe" DO NOT USE FOR: Zod/Yup projects (use zod/yup skills), non-decorator patterns, client-side forms (prefer Zod), non-NestJS backends

Deno testing with Deno.test and std/testing. Covers unit tests, async tests, mocking, snapshots, and BDD-style testing. USE WHEN: user mentions "deno test", "deno testing", asks about "Deno.test", "assertSnapshot", "FakeTime", "spy", "stub", "BDD deno" DO NOT USE FOR: Node.js projects - use `vitest` or `jest`; Browser-only code - use Playwright; Python - use `pytest`; Java - use `junit`; Non-Deno TypeScript - use `vitest`

Elasticsearch search and analytics engine. Full-text search, aggregations, document store. Use when implementing search functionality or log analytics. USE WHEN: user mentions "elasticsearch", "full-text search", "search indexing", "log analytics", "ELK stack", "aggregations", "faceted search", "autocomplete", "suggestions" DO NOT USE FOR: primary database - use `postgresql` or `mongodb` instead, caching - use `redis` instead, ACID transactions - use SQL databases instead

Email ingestion for RAG. Covers EML (Python email module, mailparser), MSG (extract-msg), Gmail and Outlook APIs, thread reconstruction via In-Reply-To / References, recursive attachment handling, signature/quote redaction, chain deduplication, and metadata (sender, date, subject) preservation. USE WHEN: user mentions "email ingestion", "EML", "MSG file", "mailparser", "extract-msg", "Gmail API", "Outlook API", "Microsoft Graph mail", "email thread", "reply-to header", "email attachments", "signature stripping" DO NOT USE FOR: PDF attachments inside emails (extract here, then route to `pdf-extraction`); office attachments (route to `office-docs`); image attachments needing OCR - route to `ocr`; calendar ICS files - not covered here

Color palettes for 2D games: color theory practical (warm/cool, complementary, analogous), restricted palettes ready-made (PICO-8, GameBoy DMG, DB16, DB32, AAP-64, NES, C64), hue shifting (warm-cool ramps), palette swaps for character variants and time-of-day, indexed mode workflow, .gpl/.pal/.ase/.json export. USE WHEN: starting a project's palette, choosing restricted palette, designing color ramps for shading, planning palette swaps.

Go programming language. Covers goroutines, channels, interfaces, error handling, and modules. Use for building concurrent, high-performance backend services. USE WHEN: user mentions "go", "golang", "goroutines", "channels", asks about "concurrency", "select statement", "interfaces", "error handling", "go modules" DO NOT USE FOR: Gin/Fiber/Echo frameworks - use framework-specific skills DO NOT USE FOR: GORM - use ORM-specific skill DO NOT USE FOR: gRPC - use API design skills

Java logging with SLF4J facade, Logback, and Log4j2 implementations. Covers configuration, log levels, structured logging, async logging, and production best practices for Spring Boot applications. USE WHEN: user mentions "java logging", "spring boot logging", "slf4j setup", asks about "how to log in java", "logback vs log4j2", "java logging best practices" DO NOT USE FOR: Node.js logging - use `nodejs-logging` instead, Python logging - use `python-logging`, Kotlin-specific logging - similar but has nuances

Jest testing framework. Covers unit tests, mocking, and snapshots. Use for testing React and Node.js projects. USE WHEN: user mentions "jest", "jest test", "snapshot test", asks about "jest.fn", "jest.mock", "toMatchSnapshot", "React testing" DO NOT USE FOR: E2E tests - use `playwright` or `cypress`; Vite projects - use `vitest` instead; React component testing alone - combine with `testing-library`

Structured markup ingestion for RAG. Covers markdown (markdown-it, tree-sitter), AsciiDoc, reStructuredText, frontmatter (YAML/TOML), header-hierarchy metadata, code-block-aware chunking (never split fenced blocks), tables, internal link resolution, Obsidian/Dendron vaults, and MDX component stripping. USE WHEN: user mentions "markdown chunking", "header splitter", "frontmatter", "code-block chunking", "Obsidian vault", "Dendron", "MDX", "AsciiDoc", "reStructuredText", "markdown-it", "tree-sitter markdown", "MarkdownHeaderTextSplitter" DO NOT USE FOR: HTML from the web - use `web-scraping`; office docs - use `office-docs`; PDF-rendered markdown-looking content - use `pdf-extraction`; generic partitioning of mixed filetypes - use `unstructured-io`

Integration testing patterns for messaging brokers: Redis Pub/Sub, NATS, Pulsar, SQS, ActiveMQ, Azure Service Bus, and Google Pub/Sub. Container-based and emulator-based testing for Java, Node.js, and Python. USE WHEN: user mentions "messaging test", "redis pubsub test", "nats test", "pulsar test", "sqs test", "localstack test", "activemq test", "azure service bus test", "google pubsub test", "message queue test" DO NOT USE FOR: Kafka testing - use `messaging-testing-kafka`; RabbitMQ testing - use `messaging-testing-rabbitmq`; Generic testcontainers - use `testcontainers` skill

MongoDB document database. Covers CRUD operations, aggregations, and indexes. Use when working with MongoDB. USE WHEN: user mentions "mongodb", "mongo", "document database", asks about "aggregation pipeline", "embedded documents", "BSON", "replica sets", "sharding", "atlas" DO NOT USE FOR: SQL databases - use `postgresql`/`mysql` instead, Redis - use `redis` instead, Elasticsearch - use `elasticsearch` instead, Relational data - use SQL skills

Office document ingestion for RAG: DOCX (python-docx, mammoth to markdown), PPTX (python-pptx with image extraction), XLSX (openpyxl, pandas, table-aware chunking), plus Notion/Confluence/Quip export to markdown. Preserves headings, lists, comments, and embedded images. USE WHEN: user mentions "DOCX", "Word document", "python-docx", "mammoth", "PPTX", "PowerPoint", "python-pptx", "XLSX", "Excel", "openpyxl", "spreadsheet", "Notion export", "Confluence export", "Quip export", "office documents" DO NOT USE FOR: generic multi-format partitioning - use `unstructured-io`; PDF exports of office docs - use `pdf-extraction`; structured markdown vaults - use `markdown-structured`; emails with office attachments (handle extraction here, ingestion via `email-ingestion`)

OSV-Scanner — language-agnostic vulnerability scanner by Google, queries the OSV.dev database (aggregates RustSec, GHSA, PyPA, npm advisories, Go vulndb, etc.). Scans Cargo.lock, package-lock.json, pnpm-lock.yaml, Pipfile.lock, go.sum, Gradle dependency tree, container images, and Git submodules. Replaces multiple language-specific scanners with one tool. CI integration with SARIF output. USE WHEN: user mentions "osv-scanner", "osv.dev", "OSV scanner", "OSV database", "GHSA", "language-agnostic vuln scan", "supply chain vulnerability scan", "deps.dev", "Google OSV" DO NOT USE FOR: Rust-specific advisory enforcement (policy) - use `quality/rust-supply-chain` DO NOT USE FOR: Container scanning depth - use Trivy / Grype skills DO NOT USE FOR: Code SAST (CodeQL, Semgrep) - use security-specific skills

PostgreSQL relational database. Covers SQL queries, indexes, constraints, and performance. Use when working with PostgreSQL. USE WHEN: user mentions "postgres", "postgresql", "pg_", asks about "JSONB queries", "window functions", "recursive CTE", "row level security", "full text search", "partitioning", "pgBouncer", "replication" DO NOT USE FOR: MySQL syntax - use `mysql` instead, MongoDB - use `mongodb` instead, Oracle PL/SQL - use `plsql` instead, SQL Server T-SQL - use `tsql` instead

React patterns for API consumption. Covers custom hooks, Suspense, SWR, error boundaries, and real-time updates. USE WHEN: user mentions "data fetching in React", "useFetch", "SWR", "fetch hook", "API integration", "REST API", asks about "React data loading", "custom fetch hooks" DO NOT USE FOR: TanStack Query specific features - use `state-tanstack-query`, GraphQL - use GraphQL-specific libraries, Non-React frameworks

React Server Components (RSC) for server-side rendering without client JS. Covers Server vs Client components, data fetching patterns, Server Actions, streaming, caching, and Next.js App Router integration. USE WHEN: user mentions "Server Components", "RSC", "Next.js App Router", "use server", "Server Actions", "async components", asks about "server-side React", "zero bundle impact" DO NOT USE FOR: Client-only React apps (CRA, Vite), Next.js Pages Router, React < 18, Non-Next.js frameworks without RSC support

Redis Pub/Sub and Streams for messaging. Covers publish/subscribe, streams with consumer groups, and real-time patterns. Use for lightweight messaging and real-time features. USE WHEN: user mentions "redis pub/sub", "redis streams", "real-time messaging", "lightweight messaging", "pattern subscriptions", asks about "fast messaging", "simple pub/sub", "event streams with redis" DO NOT USE FOR: guaranteed delivery - use `kafka`, `rabbitmq`, or `pulsar`; complex routing - use `rabbitmq`; high durability needs - use `kafka`; enterprise features - use `activemq`; cloud-native - use cloud providers

rust_decimal — fixed-point decimal arithmetic for Rust. 128-bit signed decimal, exact (no float rounding errors), suitable for finance, currency, Bitcoin sat ↔ fiat conversions, accounting. Up to 28 significant digits, arithmetic preserves precision. Includes RoundingStrategy for explicit truncation/rounding control. Integrates with serde, sqlx, postgres. USE WHEN: user mentions "rust_decimal", "Decimal", "dec!", "RoundingStrategy", "no float", "exact arithmetic", "fixed-point Rust", "currency math Rust", "sat to fiat conversion" DO NOT USE FOR: Crypto primitives - use `security/libsodium` or `bitcoin/cryptography/*` DO NOT USE FOR: BigInteger / arbitrary-precision math - use `num-bigint` DO NOT USE FOR: Time/duration math - use `chrono` or `time` crate

Spring Data Elasticsearch for full-text search and analytics. Covers ElasticsearchOperations, repositories, aggregations, and index management. USE WHEN: user mentions "spring data elasticsearch", "ElasticsearchRepository", "ElasticsearchOperations", "@Document elasticsearch", "Spring Boot Elasticsearch" DO NOT USE FOR: raw Elasticsearch queries - use `elasticsearch` instead, ELK stack setup - use `elasticsearch` instead

Turbine — small Kotlin testing library for kotlinx.coroutines Flows. Provides ergonomic API to test Flow emissions deterministically: awaitItem, expectMostRecentItem, awaitComplete, awaitError. Works with StateFlow, SharedFlow, Channel-backed flows, combine/map/debounce. KMP-friendly. USE WHEN: user mentions "Turbine", "app.cash.turbine", ".test {}", "awaitItem", "Flow testing", "StateFlow test", "SharedFlow test", "expectMostRecentItem", "cancelAndIgnoreRemainingEvents" DO NOT USE FOR: Mobile E2E - use `testing/maestro` DO NOT USE FOR: Compose snapshot - use `testing/compose-snapshot` DO NOT USE FOR: Generic Kotlin testing - use `testing/kotest` DO NOT USE FOR: Suspend function (non-Flow) testing - use `kotlinx-coroutines-test` directly

WireGuard VPN setup on Linux. Covers server config, client config, key management, NAT/IP forwarding, peer management, split tunnel, full tunnel, and site-to-site. USE WHEN: user mentions "wireguard", "wg0", "wg-quick", "wireguard vpn", "wireguard server setup", "wireguard peer", "wireguard keys", "wireguard client", "wireguard nat", "wireguard iptables", "wireguard split tunnel", "wireguard qr code", "wg genkey", "wireguard site-to-site", "wireguard keepalive" DO NOT USE FOR: OpenVPN setups - different tool and protocol, IPsec / IKEv2 VPNs, Tailscale or Netbird (they use WireGuard underneath but have their own CLIs), Cloud VPN gateways (AWS VPN, GCP Cloud VPN)

BIP340 Schnorr signatures over secp256k1: sign/verify, x-only pubkeys, tagged hashes, batch verification, key tweaking. Building block for Taproot, MuSig2, FROST, adaptor signatures, DLCs. Quick refs: signing pseudocode, batch verification, common mistakes. USE WHEN: signing/verifying Taproot key-path or Tapscript spends, designing aggregation protocols, reading BIP340.

Anthropic's Contextual Retrieval technique in depth. Prepend LLM-generated chunk-specific context (Claude Haiku) to each chunk before indexing. Combines contextual BM25 + contextual embeddings + reranking for up to 67% retrieval failure reduction. Full production pipeline with prompt caching (90% cost cut), batch processing, and eval numbers. USE WHEN: user mentions "contextual retrieval", "contextual embeddings", "Anthropic contextual retrieval", "chunk context", "contextual BM25", "49% retrieval improvement" DO NOT USE FOR: generic chunking - use `chunking-strategies`; hybrid search fundamentals - use `hybrid-search`; reranking on its own - use `reranking`

Deno runtime for TypeScript/JavaScript. Covers permissions, standard library, testing, and Deploy. Use for secure, TypeScript-native backend development. USE WHEN: user mentions "deno", "permissions", "Deno.serve", asks about "deno test", "deno deploy", "standard library", "top-level await", "npm compatibility" DO NOT USE FOR: Node.js runtime - use `nodejs` skill instead DO NOT USE FOR: Fresh/Oak frameworks - use framework-specific skills DO NOT USE FOR: Language syntax - use `typescript` skill

Interaction design patterns for web interfaces. Covers motion and animation timing, microinteractions, loading states, skeleton screens, optimistic UI, form UX, mobile touch targets, thumb zones, and responsive interaction patterns. USE WHEN: user mentions "animation", "transition", "loading state", "skeleton screen", "microinteraction", "form UX", "touch target", "mobile UX", "thumb zone", "optimistic UI", "perceived performance", "CLS", "reduced motion", "bottom navigation", "hamburger menu", asks "how fast should animations be" DO NOT USE FOR: CSS animation syntax (use styling/tailwindcss), performance profiling (use best-practices/performance), WCAG audit (use accessibility/wcag)

Building knowledge graphs from unstructured text. LLM-based triple extraction (subject-predicate-object), schema-guided extraction via Pydantic + structured output, REBEL model, OpenIE, entity linking to Wikidata/DBpedia, validation with LLM judges, incremental KG updates, and exporting to Neo4j / Amazon Neptune / TigerGraph. Full pipeline. USE WHEN: user mentions "knowledge graph construction", "KG construction", "triple extraction", "OpenIE", "REBEL", "Pydantic triples", "entity linking Wikidata", "build knowledge graph", "extract triples" DO NOT USE FOR: graph RAG retrieval - use `graph-rag`; deduping/canonicalizing entities - use `entity-resolution`; ontology-aware retrieval - use `ontology-guided-retrieval`

Kotlin code quality toolchain — detekt (static analysis with rule sets, baseline, custom rules), ktlint (formatter + linter for the Kotlin style guide), and Compose-specific lint rules. Covers Gradle integration (KMP-aware), pre-commit hooks, CI integration, baseline files for legacy code, and config patterns for production apps including Jetpack Compose specifics (detekt-compose-rules). USE WHEN: user mentions "detekt", "ktlint", "ktlint-gradle", "detekt-gradle", "kotlin static analysis", "kotlin formatter", "kotlin lint", "compose-rules detekt", "kotlin code style", "detekt baseline" DO NOT USE FOR: Java-only static analysis - use SpotBugs/Checkstyle skill DO NOT USE FOR: Rust code quality - use `quality/rust-supply-chain` DO NOT USE FOR: Security-specific Kotlin issues - use `security/kotlin-security` DO NOT USE FOR: TypeScript linting - use eslint-specific skill

Creates MCP (Model Context Protocol) servers for Claude Code. Covers TypeScript implementation with MCP SDK, tool definitions, transport types, metadata.json for dev-suite, and deployment patterns. Follows official MCP specification. USE WHEN: user mentions "MCP server", "create MCP", "model context protocol", "MCP tool", "external tool", "mcp-servers/", "metadata.json", "tool integration" DO NOT USE FOR: configuring existing MCP servers - check MCP docs; creating skills - use `skill-authoring`; creating hooks - use `hook-authoring`

Python logging with the standard library logging module and structlog. Covers log levels, handlers, formatters, structured logging, and production best practices for FastAPI/Django applications. USE WHEN: user mentions "python logging", "fastapi logging", "django logging", asks about "how to log in python", "python logging module", "logging configuration python" DO NOT USE FOR: Node.js logging - use `nodejs-logging` instead, Java logging - use `slf4j` or `logback`, structlog-specific - use `structlog` skill for deep dive

React Hooks patterns, rules, and custom hook development. Covers all built-in hooks (useState, useEffect, useContext, etc.), performance hooks, and patterns for creating reusable custom hooks. USE WHEN: user mentions "custom hooks", "hook rules", "useState", "useEffect", "useReducer", "useCallback", "useMemo", asks about "hook dependencies", "creating hooks", "hook patterns" DO NOT USE FOR: React 19 use() hook - use `react-19` skill instead, Context API details - use `react-context` skill instead, form-specific hooks - use `react-forms` or `react-hook-form` skills instead

Creates and improves Claude Code Skills (SKILL.md files). Covers frontmatter, descriptions, progressive disclosure, bundled resources, and content guidelines. Follows official Anthropic best practices. USE WHEN: user mentions "create skill", "write skill", "SKILL.md", "skill file", "slash command", "custom command", "extend Claude", "skill description", "skill not triggering", "skill authoring" DO NOT USE FOR: creating agents/subagents - use `agent-authoring`; creating hooks - use `hook-authoring`; creating MCP servers - use `mcp-authoring`; creating plugins - use `plugin-authoring`

Video ingestion for multimodal RAG. Covers keyframe extraction (ffmpeg, PySceneDetect), audio-track transcription via Whisper, visual embeddings per keyframe (CLIP, SigLIP, VoyageAI multimodal), multi-modal retrieval combining visual + transcript hits, timestamp-anchored citations, and YouTube transcript ingestion. USE WHEN: user mentions "video RAG", "video transcription", "keyframe extraction", "PySceneDetect", "ffmpeg scenes", "CLIP embeddings", "SigLIP", "VoyageAI multimodal", "YouTube transcript", "multimodal retrieval", "video chunking" DO NOT USE FOR: audio-only content (podcasts, meetings) - use `audio-transcription`; screen recordings that are effectively slide decks - use `office-docs` on the source PPTX; static images - use `ocr` or pure vision embeddings

Audio transcription pipeline for RAG. Covers OpenAI Whisper (local via whisper.cpp, faster-whisper, Groq API), AssemblyAI (diarization, sentiment), Deepgram Nova-3 (real-time), timestamped chunking, speaker-aware chunking, multilingual (Whisper large-v3), and embedding transcripts for retrieval. USE WHEN: user mentions "transcribe audio", "Whisper", "whisper.cpp", "faster-whisper", "Groq Whisper", "AssemblyAI", "Deepgram", "speaker diarization", "podcast transcription", "meeting transcript", "multilingual audio" DO NOT USE FOR: video with visuals - use `video-rag`; live voice assistants (streaming to LLM) - not a RAG concern; audio that is actually a phone voicemail attachment in email - handle via `email-ingestion` then reuse this skill

Bitcoin Script language: stack execution model, opcodes, all output types (P2PK, P2PKH, P2SH, P2WPKH, P2WSH, P2TR, Tapscript), witness construction, sig encoding (DER/Schnorr). Quick refs: opcode reference, output-type construction, Tapscript diff vs legacy. USE WHEN: writing/decoding scriptPubKey, building scriptSig/witness, understanding "what does this script do", reading raw txs.

Multi-turn RAG: chat history management, context window compaction (summarization, sliding window, vector memory), query rewriting with coreference resolution, follow-up vs new-query routing, LangChain ConversationalRetrievalChain, LlamaIndex ChatEngine, Redis/Postgres message history stores. USE WHEN: user mentions "conversational RAG", "multi-turn RAG", "chat history", "follow-up question", "chat memory", "ChatEngine", "ConversationalRetrievalChain", "coreference in RAG" DO NOT USE FOR: single-turn pipelines - use `rag-architecture`; query rewriting in isolation - use `query-transformations`; user-level long-term memory - use `personalization-rag`

Fine-tuning cross-encoders for domain-specific reranking. Training data (query-doc relevance labels, MS MARCO format), sentence-transformers CrossEncoder API, loss functions (BCE, margin), hard negative mining from BM25 and dense retrievers, distillation from strong teacher rerankers (BGE-reranker-v2, Cohere) into small models, NDCG@10 evaluation. USE WHEN: user mentions "fine-tune cross-encoder", "train reranker", "hard negative mining", "MS MARCO triples", "knowledge distillation reranker", "CrossEncoder", "cross-encoder training" DO NOT USE FOR: zero-shot reranking with existing APIs - use `rag/reranking`; LLM reranker - use `retrieval/rank-gpt`; ColBERT training - use `retrieval/colbert-retrieval`; SPLADE training - use `retrieval/splade-deep`

.NET and ASP.NET Core security patterns. Covers Identity, authentication, dependency auditing, secure coding practices, and OWASP for .NET ecosystem. USE WHEN: user works with "C#", ".NET", "ASP.NET Core", "Entity Framework", asks about ".NET vulnerabilities", "NuGet security", ".NET authentication", "Blazor security" DO NOT USE FOR: general OWASP concepts - use `owasp` or `owasp-top-10` instead, Java/Python security - use language-specific skills

Entity resolution (ER) for multi-source RAG and knowledge graphs. Dedup entities across documents ("Acme Corp" = "Acme, Inc." = "Acme Corporation"). Covers blocking, probabilistic linking (Splink), Dedupe.io, embedding-based matching, LLM-assisted resolution with rules, graph-based resolution (merged nodes), canonical IDs, and mapping tables. USE WHEN: user mentions "entity resolution", "record linkage", "dedup", "deduplication", "Splink", "Dedupe.io", "canonicalization", "entity matching", "coreference resolution" (corpus-level) DO NOT USE FOR: building a KG from scratch - use `knowledge-graph-construction`; retrieval tactics - use `graph-rag`; ontology-aware retrieval - use `ontology-guided-retrieval`

Fiber Go web framework inspired by Express. Covers routing, middleware, context, WebSocket, and prefork. Use for Express-like Go development. USE WHEN: user mentions "fiber", "gofiber", "express-like go", "fasthttp go", asks about "fiber middleware", "fiber context", "fiber prefork", "fiber websocket", "go express", "fast go framework" DO NOT USE FOR: Gin projects - use `gin` instead, Echo projects - use `echo` instead, Chi projects - use `chi` instead, standard net/http compatibility required

Production ingestion pipelines with Airflow, Prefect 3, Dagster. DAG design for RAG: extract -> parse -> chunk -> embed -> index. Retry policies, idempotency, partial failure, monitoring, backfills, incremental vs full refresh, data lineage, upstream dependencies. Full Dagster and Prefect examples. USE WHEN: user mentions "ingestion pipeline", "Airflow RAG", "Prefect RAG", "Dagster RAG", "DAG for embeddings", "backfill embeddings", "incremental ingestion", "idempotent ingestion" DO NOT USE FOR: real-time streaming ingestion - use `cdc-streaming-ingestion`; chunking details - use `chunking-strategies`; eval - use `rag-evaluation`

Java code quality with Checkstyle, SpotBugs, PMD, and SonarJava. Covers static analysis, code style, and best practices. USE WHEN: user works with "Java", "Spring Boot", "Maven", "Gradle", asks about "Checkstyle", "SpotBugs", "PMD", "Java code smells", "Java best practices" DO NOT USE FOR: SonarQube generic - use `sonarqube` skill, testing - use Spring Boot test skills, security - use `java-security` skill

OpenTelemetry - vendor-neutral observability framework for distributed systems. Provides traces, metrics, and logs with standardized instrumentation and exporters. USE WHEN: user mentions "opentelemetry", "otel", "distributed tracing", "observability", asks about "how to trace microservices", "opentelemetry setup", "jaeger integration", "prometheus metrics" DO NOT USE FOR: Application logging only - use logging skills instead, APM vendor-specific - use vendor docs, Simple monitoring - Prometheus/Grafana may be sufficient

Creates Claude Code plugins that package skills, agents, hooks, and MCP servers into distributable bundles. Covers plugin.json manifest, directory structure, namespacing, and distribution. Follows official Anthropic best practices. USE WHEN: user mentions "create plugin", "plugin", "plugin.json", "distribute skills", "share agents", "plugin marketplace", "package extension", ".claude-plugin" DO NOT USE FOR: creating standalone skills - use `skill-authoring`; creating standalone agents - use `agent-authoring`; creating standalone hooks - use `hook-authoring`

RabbitMQ message broker with AMQP protocol. Covers exchanges, queues, bindings, and messaging patterns. Use for reliable message delivery and complex routing scenarios. USE WHEN: user mentions "rabbitmq", "amqp", "exchanges", "routing patterns", "topic exchange", "fanout", asks about "message routing", "work queues", "request/reply", "flexible routing" DO NOT USE FOR: high-throughput streaming - use `kafka` or `pulsar`; cloud-native - use `nats`; AWS-native - use `sqs`; JMS required - use `activemq`; simple pub/sub - use `redis-pubsub`

Security controls for RAG. Indirect prompt-injection via retrieved documents, PII detection/redaction (Microsoft Presidio, AWS Comprehend), multi-tenant isolation, ACL-aware retrieval with row-level/metadata filtering, data-leakage prevention, jailbreak hardening on retrieved context, GDPR right-to-be-forgotten in vector DBs. USE WHEN: user mentions "prompt injection RAG", "indirect prompt injection", "PII redaction", "Presidio", "ACL RAG", "row-level security", "multi-tenant RAG isolation", "GDPR vector DB", "right to be forgotten", "jailbreak", "data leakage RAG" DO NOT USE FOR: hallucination detection - use `rag-guardrails`; tenancy scaling patterns - use `rag-production`; audit tracing schema - use `rag-observability`

React 18+ library for building user interfaces. Covers components, hooks, state management, and rendering patterns. USE WHEN: user mentions "React component", "useState", "useEffect", "hooks", asks about "building UI", "component lifecycle", "React rendering", "JSX" DO NOT USE FOR: React 19 features - use `react-19` instead, React Router - use `react-router` instead, performance optimization - use `react-performance` instead

Rust security patterns for web applications. Covers memory safety guarantees, dependency auditing, secure coding practices, and OWASP for Rust ecosystem. USE WHEN: user works with "Rust", "Actix", "Axum", "Rocket", "Warp", asks about "Rust vulnerabilities", "cargo audit", "Rust injection", "Rust authentication" DO NOT USE FOR: general OWASP concepts - use `owasp` or `owasp-top-10` instead, other language security - use language-specific skills

LangChain SelfQueryRetriever pattern. LLM infers structured metadata filters from natural language ("books by Asimov after 2000" -> filter author=Asimov AND year>2000). Metadata schema declaration, comparators and operators, LlamaIndex AutoRetriever equivalent, combining with hybrid search, evaluation of filter correctness. USE WHEN: user mentions "self-querying retriever", "SelfQueryRetriever", "auto retriever", "metadata filter from query", "NL to filter", "AutoRetriever" DO NOT USE FOR: text-to-SQL on tables - use `tabular-rag`; plain query rewriting - use `query-transformations`; hybrid search - use `hybrid-search`

Spring AMQP for RabbitMQ integration with RabbitTemplate and @RabbitListener. Covers exchanges, queues, bindings, DLQ, and message conversion. USE WHEN: user mentions "spring amqp", "RabbitTemplate", "@RabbitListener", "RabbitMQ Spring", "AMQP", "message queue Spring Boot" DO NOT USE FOR: raw RabbitMQ - use `rabbitmq` skill, Kafka - use `spring-kafka` skill, generic messaging patterns - use `messaging-expert`

Spring Authorization Server for building OAuth 2.1 and OpenID Connect providers. USE WHEN: user mentions "OAuth2 server", "authorization server", "OIDC provider", "token endpoint", asks about "how to implement OAuth2", "create authorization server", "issue JWT tokens", "custom OAuth provider" DO NOT USE FOR: OAuth2 client configuration - use `spring-security` instead, resource server JWT validation - use `spring-security` instead

Spring Data Redis for caching, session storage, and data persistence. Covers RedisTemplate, @Cacheable, repositories, pub/sub, and distributed locks. USE WHEN: user mentions "spring data redis", "RedisTemplate", "@Cacheable", "Spring Boot caching", "@RedisHash", "spring session redis", "distributed lock Spring" DO NOT USE FOR: raw Redis commands - use `redis` instead, non-Spring Redis clients - use `redis` instead

Spring HATEOAS for building hypermedia-driven RESTful APIs. Covers EntityModel, CollectionModel, RepresentationModelAssembler, HAL, and affordances. USE WHEN: user mentions "HATEOAS", "hypermedia", "HAL", "EntityModel", "CollectionModel", "RepresentationModelAssembler", "Richardson Maturity Level 3" DO NOT USE FOR: simple REST APIs without hypermedia - use plain Spring MVC, GraphQL - use `spring-graphql` skill

Spring Mail for sending emails in Spring Boot 3.x. Covers JavaMailSender, SimpleMailMessage, MimeMessage, HTML emails, attachments, templates (Thymeleaf/FreeMarker), async sending, and testing. USE WHEN: user mentions "spring mail", "JavaMailSender", "email Spring Boot", "send email", "MimeMessage", "email templates", "SMTP Spring" DO NOT USE FOR: transactional email services (SendGrid, SES) - use their SDKs, SMS/push notifications - use appropriate services

Web scraping for RAG ingestion. Covers Firecrawl API, Crawl4AI (open source), trafilatura (main-content extraction), BeautifulSoup + readability-lxml, Scrapy for large crawls, Playwright for JS-rendered pages, sitemap.xml discovery, robots.txt, polite crawling, anti-bot, and deduplication. USE WHEN: user mentions "web scraping", "crawl website", "Firecrawl", "Crawl4AI", "trafilatura", "readability", "BeautifulSoup", "Scrapy", "Playwright scrape", "sitemap", "robots.txt", "crawl for RAG", "extract article" DO NOT USE FOR: parsing already-downloaded HTML files for layout - use `unstructured-io`; PDF downloads from the web - use `pdf-extraction`; office docs behind login - use `office-docs`; structured markdown vaults - use `markdown-structured`

Retrieval strategies beyond top-K similarity. Parent-document, small-to-big, multi-vector, contextual compression, sentence-window, auto-merging, RAPTOR, and hierarchical indexing. LangChain + LlamaIndex code with tradeoffs. USE WHEN: user mentions "parent document retriever", "small-to-big", "multi-vector", "sentence window", "auto-merging", "RAPTOR", "hierarchical index", "contextual compression" DO NOT USE FOR: chunking the source docs - use `chunking-strategies`; query rewriting - use `query-transformations`; BM25+dense fusion - use `hybrid-search`; reranker models - use `reranking`

Production RAG templates for five domains: customer-support, developer-docs, legal, medical, financial. Each covers corpus shape, chunking, embedding model choice, metadata schema, domain-specific guardrails (hallucination tolerance, PII, compliance), and evaluation criteria. USE WHEN: user mentions "RAG for support", "legal RAG", "medical RAG", "developer docs RAG", "financial RAG", "domain template RAG" DO NOT USE FOR: general architecture - use `rag-architecture`; evaluation methodology - use `rag-evaluation`; security specifics - use `rag-security`

User feedback signals for RAG improvement: thumbs up/down, click-through, dwell time, explicit ratings. Implicit vs explicit signals, logging schema, feedback -> retraining pipelines (embedding fine-tuning with hard negatives, reranker fine-tuning from CTR), A/B testing RAG variants, Langfuse/LangSmith feedback APIs, feature stores for online learning. USE WHEN: user mentions "user feedback", "thumbs up down", "CTR", "dwell time", "RAG evaluation feedback", "hard negatives", "reranker fine-tune", "A/B test RAG" DO NOT USE FOR: offline eval metrics - use `rag-evaluation`; retriever algorithms - use `advanced-retrieval`; observability as such - use `rag-observability`

AI-assisted 2D / pixel-art generation tools and services. Covers pixel-art-specific generators (PixelLab, Retro Diffusion), game-asset generators (Scenario, Leonardo.AI), DIY Stable Diffusion + LoRAs, and general models with pixel-art prompting (Midjourney, DALL-E, Imagen). Trade-off framework: control vs. speed, palette consistency, animation coherence, tile-ability, commercial license, runs locally vs. cloud, cost per asset / per month. USE WHEN: deciding which AI tool fits a project, generating placeholder assets, generating final assets at scale, integrating an AI sprite pipeline, evaluating commercial license, comparing API costs. DO NOT USE FOR: hand-painted-only workflows (use `tools` skill), 3D / model generation (out of scope), engine integration of finished sprites (use `tools` + engine-specific skill).

Go security patterns for web applications. Covers dependency auditing, secure coding practices, crypto, and OWASP for Go ecosystem. USE WHEN: user works with "Go", "Golang", "Gin", "Fiber", "Echo", asks about "Go vulnerabilities", "Go modules security", "Go injection", "Go authentication" DO NOT USE FOR: general OWASP concepts - use `owasp` or `owasp-top-10` instead, other language security - use language-specific skills

Knowledge-graph-augmented retrieval. Entity and triple extraction, graph construction (Neo4j, LlamaIndex PropertyGraphIndex), hierarchical community summarization (Microsoft GraphRAG), personalized PageRank (HippoRAG), multi-hop traversal retrieval, and hybrid graph + vector pipelines. USE WHEN: user mentions "GraphRAG", "HippoRAG", "knowledge graph RAG", "entity extraction", "multi-hop reasoning", "Neo4j RAG", "LlamaIndex property graph", "LangChain graph retriever", "triple extraction", "community summarization" DO NOT USE FOR: vanilla vector RAG - use `rag-patterns`; multimodal inputs - use `multimodal-rag`; production indexing ops - use `rag-production`; hallucination checks - use `rag-guardrails`

Windows Human Interface Device (HID) stack and filter drivers. Covers HID class architecture, transport drivers (HIDUSB / HIDBTH / HIDI2C), HIDClass, mouse/keyboard filter (Mouclass / Kbdclass) drivers, internal IOCTLs (IOCTL_HID_*), filter placement (upper/lower), event interception, injection, and the Virtual HID Framework (VHF). USE WHEN: user mentions "HID", "HID filter", "mouse filter", "keyboard filter", "touch filter", "pen filter", "HIDClass", "mouclass", "kbdclass", "moufiltr", "kbfiltr", "IOCTL_HID_READ_REPORT", "IOCTL_HID_GET_FEATURE", "VHF", "Vhf.sys", "input intercept", "input injection" DO NOT USE FOR: GPU input redirection (different stack), DirectInput/XInput (user-mode game APIs)

Windows Indirect Display Driver (IDD) framework. UMDF v2-based driver model for adding virtual monitors that the Desktop Window Manager treats as real displays, then capturing the rendered frames in user mode and routing them somewhere else (network, file, encoder) instead of a physical panel. Covers IddCx callbacks, monitor descriptors (EDID/MCCS), swap-chain processing, cursor handling, hot-plug, hardware cursor, multi-monitor, HDR, and integration with NVENC / Rivermax / RDMA NICs for low-latency network streaming. USE WHEN: user mentions "IDD", "Indirect Display", "IddCx", "IddSampleDriver", "virtual monitor", "virtual display driver", "IDDCX_SWAPCHAIN", "AcquireBuffer", "FinishedProcessingFrame", "EDID for virtual display", "stream desktop", "network display" DO NOT USE FOR: WDDM display miniport drivers (real GPU), DXGK kernel work, Miracast (different framework), Remote Desktop / RDP (different stack)

BOLT (Basis of Lightning Technology) specification index. Wire format, channel management, onion routing, gossip, invoices, feature bits. Quick refs: BOLT-by-BOLT summary, message-type tables, feature-bit registry. USE WHEN: implementing/reading Lightning protocol, looking up message types, debugging interop between implementations.

Node.js runtime best practices. Covers event loop, async patterns, streams, worker threads, memory management, and production optimization. USE WHEN: user mentions "node.js", "event loop", "streams", "worker threads", asks about "process.nextTick", "memory leaks", "cluster mode", "async patterns" DO NOT USE FOR: Express/NestJS frameworks - use framework-specific skills DO NOT USE FOR: Language syntax - use `javascript` or `typescript` skills DO NOT USE FOR: Package management - use npm/pnpm/yarn skills

Node.js logging with Winston, Pino, and built-in console. Covers structured logging, log levels, transports, async logging, and production best practices for Express/NestJS applications. USE WHEN: user mentions "node.js logging", "express logging", "nestjs logging", asks about "how to log in node", "winston vs pino", "node logging best practices" DO NOT USE FOR: Python logging - use `python-logging` instead, Java logging - use `slf4j` or `logback`, Browser logging - different environment and APIs

User-specific retrieval. Per-user namespaces/filters, preference embeddings, collaborative signals, reranking with user context (role, history, favorites), privacy-preserving design (encrypted metadata, differential privacy), GDPR- compliant personalization, long-term user memory with mem0/Zep/Letta, graph- based user-entity memory. USE WHEN: user mentions "personalized RAG", "user-specific RAG", "per-user retrieval", "mem0", "Zep", "Letta", "long-term memory", "user preferences RAG" DO NOT USE FOR: per-session chat memory - use `conversational-rag`; feedback-driven model updates - use `feedback-loops`; multi-tenant isolation only - use `rag-architecture`

Operational concerns for RAG at scale. Incremental indexing (CDC, doc diffing), blue-green re-indexing, index aliases/versioning, embedding-model hot-swap, cost optimization (dimension reduction, cheaper embed models), capacity and token budgets, p50/p95/p99 SLAs, horizontal scaling, multi-tenant isolation, async ingestion (Kafka/Pub-Sub). USE WHEN: user mentions "incremental indexing", "reindex", "index alias", "blue-green reindex", "RAG cost optimization", "multi-tenant RAG", "RAG capacity planning", "async ingestion", "CDC for RAG", "embedding model swap" DO NOT USE FOR: caching - use `rag-caching`; tracing/metrics - use `rag-observability`; access control / PII - use `rag-security`

React Concurrent Features for responsive UIs. Covers useTransition, useDeferredValue, Suspense, startTransition, concurrent rendering patterns, and priority-based updates. USE WHEN: user mentions "useTransition", "useDeferredValue", "startTransition", "concurrent rendering", "non-urgent updates", asks about "keeping UI responsive", "priority updates in React" DO NOT USE FOR: React 17 and earlier (concurrent mode not available), Simple applications without heavy rendering, Non-React frameworks

Rocket Rust web framework. Covers routing, fairings, guards, state, and testing. Use for type-safe, boilerplate-free Rust APIs. USE WHEN: user mentions "rocket", "rust type-safe api", "rocket guards", asks about "rocket fairings", "rocket state", "compile-time route checking", "rust web macros", "rocket testing" DO NOT USE FOR: Axum projects - use `axum` instead, Actix-web projects - use `actix-web` instead, Warp projects - use `warp` instead, non-Rust backends

Resilience patterns with Spring Cloud Circuit Breaker and Resilience4j. Covers circuit breaker, retry, rate limiter, bulkhead, and fallback patterns. USE WHEN: user mentions "circuit breaker", "resilience4j", "fallback", "retry pattern", "rate limiter", "bulkhead", "fault tolerance Spring" DO NOT USE FOR: basic HTTP errors - handle in code, Hystrix (deprecated) - use Resilience4j instead

Spring Data Neo4j for graph database operations. Covers node/relationship entities, Cypher queries, and Neo4jTemplate. USE WHEN: user mentions "spring data neo4j", "Neo4jRepository", "@Node", "@Relationship", "Cypher Spring", "graph database Spring Boot" DO NOT USE FOR: raw Cypher queries - consult Neo4j documentation, relational databases - use `spring-data-jpa` instead

Spring Retry for transparent retry support in Spring applications. Covers @Retryable, @Recover, RetryTemplate, backoff policies, and circuit breakers. USE WHEN: user mentions "spring retry", "@Retryable", "@Recover", "RetryTemplate", "backoff policy", "retry transient failure", "automatic retry" DO NOT USE FOR: circuit breaker patterns - use `spring-cloud-circuitbreaker` skill, message retry - use `spring-kafka` or messaging skill retry topics

Spring Session for distributed session management with Redis, JDBC, or Hazelcast. Covers session configuration, security integration, and session events. USE WHEN: user mentions "spring session", "distributed session", "session Redis", "session JDBC", "session cluster", "session management", "@SessionScope" DO NOT USE FOR: stateless JWT auth - use `jwt` skill, OAuth2 tokens - use `oauth2` skill

Spring Shell for building interactive CLI applications. Covers @ShellComponent, @ShellMethod, input validation, tables, and command groups. USE WHEN: user mentions "spring shell", "@ShellComponent", "@ShellMethod", "CLI application Spring", "interactive shell", "command line tool Spring" DO NOT USE FOR: simple scripts - use regular main method, batch processing - use `spring-batch` skill, web endpoints - use `spring-rest` skill

Springdoc OpenAPI for API documentation in Spring Boot. Covers Swagger UI configuration, annotations, schema customization, and security documentation. Based on production patterns from castellino and gestionale-presenze projects. USE WHEN: user mentions "Springdoc", "Spring Boot OpenAPI", "Swagger in Spring", "@Operation", "@Schema", "Swagger UI Spring", asks about "Spring Boot API documentation", "Spring REST documentation", "OpenAPI in Java" DO NOT USE FOR: General OpenAPI specs - use `openapi` instead; GraphQL - use `graphql` instead; Non-Spring Boot projects; Frontend OpenAPI generation - use `openapi-codegen` instead

Streaming LLM responses with inline citations. Token-level source attribution, SSE vs WebSocket, TTFT optimization, progressive disclosure (retrieval status then tokens), Python async generators, Vercel AI SDK streaming with sources, LangChain streaming callbacks, client-side citation rendering. USE WHEN: user mentions "streaming RAG", "streaming citations", "SSE RAG", "TTFT", "progressive disclosure", "AI SDK streaming", "token streaming", "inline citations" DO NOT USE FOR: generic RAG pipelines - use `rag-architecture`; citation correctness evaluation - use `rag-evaluation`; conversational memory - use `conversational-rag`

Structured data + RAG. NL2SQL hybrid patterns (text-to-SQL then execute vs embed rows), table embedding strategies (row-level, schema-level, hybrid), semantic layer integration (Cube, dbt metrics), LangChain SQLDatabaseChain, LlamaIndex PandasQueryEngine, safe SQL execution (read-only, sandboxed), schema-aware retrieval. Full PostgreSQL + pgvector hybrid code. USE WHEN: user mentions "tabular RAG", "NL2SQL", "text to SQL", "RAG on tables", "database RAG", "SQL RAG", "semantic layer", "structured data RAG" DO NOT USE FOR: unstructured doc RAG - use `rag-architecture`; metadata filtering only - use `self-querying-retriever`; KG retrieval - use `graph-rag`

Testcontainers for Docker-based integration testing in Java. Covers database containers, messaging systems, and Spring Boot @ServiceConnection. USE WHEN: user mentions "testcontainers", "docker test", "integration test", asks about "@ServiceConnection", "PostgreSQLContainer", "MongoDBContainer", "test database", "docker compose test" DO NOT USE FOR: Unit tests - use `junit`; REST API tests - use `rest-assured`; Tests without Docker - use H2 or embedded databases; Non-Java projects - check language-specific tools

Vitest testing framework. Covers unit tests, mocking, and coverage. Use for testing Vite-based and Node.js projects. USE WHEN: user mentions "vitest", "vite test", "unit test", asks about "vi.fn", "vi.mock", "test coverage", "mock functions", "test vite project" DO NOT USE FOR: E2E tests - use `playwright` instead; React component testing - combine with `testing-library`; Jest projects - use `jest` instead

WCAG 2.2 Web Content Accessibility Guidelines. Covers conformance levels A/AA/AAA, success criteria, testing with axe-core, and common accessibility patterns. USE WHEN: user mentions "accessibility", "a11y", "WCAG", "screen reader", "ARIA", asks about "color contrast", "keyboard navigation", "accessible forms", "compliance", "ADA", "Section 508" DO NOT USE FOR: automated testing implementation - use `axe-core` instead

Actix-web Rust web framework. Covers routing, extractors, middleware, state management, and WebSocket. Use for high-performance Rust APIs. USE WHEN: user mentions "actix-web", "actix", "rust web framework", "rust api", asks about "rust async web", "actix middleware", "actix extractors", "rust websocket", "high performance rust api" DO NOT USE FOR: Axum projects - use `axum` instead, Rocket projects - use `rocket` instead, Warp projects - use `warp` instead, non-Rust backends

BM25 deep tuning. k1 and b parameters with defaults per collection, field boosts, stopwords, language-specific analyzers (Italian, French, German, non-English), stemming vs lemmatization, tokenization gotchas, Elasticsearch vs Lucene vs rank_bm25. When BM25 alone beats vectors. USE WHEN: user mentions "BM25", "BM25 tuning", "k1 b parameter", "Elasticsearch analyzer", "stemming", "lemmatization", "rank_bm25", "TF-IDF", "lexical search" DO NOT USE FOR: learned sparse - use `retrieval/splade-deep`; hybrid fusion - use `rag/hybrid-search`; dense retrieval - use `vector-stores/*`

Chi lightweight Go HTTP router. Covers routing, middleware, context, and patterns. Use for idiomatic, stdlib-compatible Go APIs. USE WHEN: user mentions "chi", "go-chi", "lightweight go router", "stdlib go router", asks about "chi middleware", "chi router", "chi context", "idiomatic go api", "net/http compatible router", "chi patterns" DO NOT USE FOR: Gin projects - use `gin` instead, Echo projects - use `echo` instead, Fiber projects - use `fiber` instead, non-Go backends

ChromaDB 0.5+ advanced features. Persistent vs ephemeral clients, collection-level HNSW tuning (space, construction_ef, M), multi-tenancy via collections, metadata filtering, embedding function plugins (OpenAI, Cohere, Voyage, custom), Chroma Cloud, recall tuning, migration patterns to production stores. USE WHEN: user mentions "Chroma", "ChromaDB", "chromadb-client", "Chroma Cloud", "Chroma collection", "chroma embedding function" DO NOT USE FOR: production-scale vector stores - use `vector-stores/qdrant-advanced`, `vector-stores/milvus`, or `vector-stores/pinecone-advanced`; hybrid search theory - use `rag/hybrid-search`

Document chunking techniques for RAG. Fixed-size, recursive, semantic, token-based, document-aware, proposition, parent-child, sliding window, and Anthropic contextual retrieval. Tradeoff tables, LangChain and LlamaIndex code. USE WHEN: user mentions "chunking", "text splitter", "split documents", "semantic chunking", "contextual retrieval", "parent-child chunks", "proposition chunking" DO NOT USE FOR: retrieval after chunking - use `advanced-retrieval`; query-side transforms - use `query-transformations`; overall design - use `rag-architecture`

.NET code quality with Roslyn analyzers, StyleCop, and dotnet format. Covers linting, formatting, and best practices for C#/.NET. USE WHEN: user works with "C#", ".NET", "ASP.NET Core", asks about "Roslyn analyzers", "StyleCop", "dotnet format", ".NET linting", "C# best practices" DO NOT USE FOR: SonarQube - use `sonarqube` skill, testing - use .NET test skills, security - use `dotnet-security` skill

Patterns for synchronizing DTOs between frontend and backend. Covers shared types, code generation, and validation sync. USE WHEN: user asks about "DTO sync", "shared types", "frontend backend types", "type consistency", "API models" DO NOT USE FOR: type generation tools - use `type-generation` skill, validation rules - use validation skills