メニュー
SkillsMP は kalilurrahman/kr-claudiator-skills から 529 件の skill を収集しています。skill を開くとソースと詳細を確認できます。
収集済み skill 529 件中 40 件を表示しています。
Cut AI/LLM spend without degrading quality — model tiering and cascade routing, Batch API offloading, context and output-token discipline, reasoning-effort tuning, and per-feature cost attribution. Outputs a per-feature cost baseline, a model routing plan…
原文の言語: 英語
Implement data observability across the five pillars — freshness, volume, schema, distribution, lineage — with anomaly detection on pipelines and lineage-driven incident triage. Outputs a tiered coverage plan, monitor configurations, alert routing rules, an…
原文の言語: 英語
Build production vision and document-understanding pipelines on the Claude API — image inputs, PDF processing, structured extraction, and multimodal RAG. Outputs ingestion-path decisions, resolution and token budgets, extraction schemas, evaluation plans, and…
原文の言語: 英語
Design and implement graph analytics — property graph vs RDF modeling, engine selection (Neo4j, Memgraph, Neptune, Spark GraphFrames), and algorithm-to-business-question mapping. Outputs graph schema, Cypher queries, batch algorithm pipelines, and GraphRAG…
原文の言語: 英語
Secure webhook receiving endpoints against spoofing, replay, duplicate processing, and overload. Outputs HMAC verification code with timing-safe compares, timestamp and idempotency defenses, secret rotation plans, queue-backed 2xx-fast handler architecture,…
原文の言語: 英語
Design and run chaos engineering experiments against infrastructure — steady-state hypotheses, blast-radius control, and staged promotion from dev to production. Outputs experiment specs (Chaos Mesh, LitmusChaos, AWS FIS), abort conditions, game-day runbooks,…
原文の言語: 英語
Design a semantic layer that defines every business metric once and serves it consistently to BI tools, spreadsheets, and AI agents. Outputs metric specs (dbt MetricFlow or Cube), a platform decision record, a governance and ownership model, and integration…
原文の言語: 英語
Build and keep alive a service catalog / internal developer portal — Backstage or a commercial IDP — with enforced ownership, on-call mapping, and maturity scorecards. Outputs catalog schema (catalog-info.yaml), ownership model, scorecard definitions, CI…
原文の言語: 英語
Run a dbt project like a software project. Outputs staging/intermediate/marts layering, naming and ref() conventions, a unit-vs-data testing strategy, incremental model configs with failure-mode defenses, documentation and exposures, and slim CI pipelines…
原文の言語: 英語
Design an identity governance and administration (IGA) program from an org's app inventory, HR system, and compliance drivers. Outputs a role and entitlement model, joiner-mover-leaver automation design, SCIM provisioning flows, an SoD ruleset, an access…
原文の言語: 英語
Design and implement streaming APIs — SSE, WebSockets, HTTP chunked transfer, and gRPC streams. Outputs a protocol selection decision, event contract, backpressure and reconnection strategy, resume-token design, auth plan for long-lived connections, and…
原文の言語: 英語
Run continuous product discovery to de-risk what you build — customer interviews, opportunity-solution trees, assumption mapping, and riskiest-assumption tests. Outputs an interview guide, an opportunity-solution tree, an assumption map with prioritized…
原文の言語: 英語
Define a product strategy that connects company vision to where-to-play and how-to-win choices, strategic bets, and quarterly execution. Outputs a strategy one-pager with diagnosis, guiding policy, positioning, bets with kill criteria, and a cascade to OKRs…
原文の言語: 英語
Design reverse ETL pipelines that sync warehouse data back into operational SaaS tools (CRM, ads, support). Outputs sync architecture, diff-based sync strategy, idempotency and rate-limit handling, identity resolution plan, and a build-vs-buy recommendation.
原文の言語: 英語
Instrument LLM applications for production observability. Outputs trace instrumentation (spans for prompts, tools, retrieval), token and cost tracking, quality monitoring with online evals, drift detection, and alerting rules aligned to OpenTelemetry GenAI…
原文の言語: 英語
Design and implement defense-in-depth guardrails for LLM applications — input/output validation, prompt-injection defense, PII redaction, topical and safety filters, and human-in-the-loop gates. Outputs a layered guard architecture, validation code,…
原文の言語: 英語
Design LLM caching across all three layers — provider-side prompt caching (Claude cache_control breakpoints and TTLs), application-level response caching, and semantic caching. Outputs cache-friendly prompt structure, breakpoint placement, TTL selection with…
原文の言語: 英語
Design a multi-channel notification system covering push/email/SMS/in-app fan-out, user preferences and quiet hours, batching and digests, idempotency, rate limiting, and delivery tracking. Outputs pipeline architecture, database schemas, worker patterns, and…
原文の言語: 英語
Design production search systems — inverted-index, vector, and hybrid retrieval, Elasticsearch/OpenSearch cluster design, indexing pipelines, and relevance tuning. Outputs index mappings, cluster topology, query pipeline design, a zero-downtime reindex…
原文の言語: 英語
Test code against thousands of generated inputs instead of hand-picked examples, letting the framework shrink every failure to a minimal counterexample. Outputs property definitions, custom generators, stateful model tests, regression pins, and CI-ready…
原文の言語: 英語
Calculate sample size and runtime for an A/B test given variance and MDE.
原文の言語: 英語
Specify tool contracts, schemas, and guardrails for an LLM agent.
原文の言語: 英語
Design multi-step Claude 5 Fable agents that plan, act, observe, and correct — with explicit budget, checkpoint, and human-in-the-loop gates you can defend in production.
原文の言語: 英語
Manages account health, relationship strategy, renewals, and expansion opportunities.
原文の言語: 英語
Acts as an agile coach who improves sprint flow, backlog hygiene, ceremonies, and delivery discipline.
原文の言語: 英語
Shapes strategy for AI-native products, differentiation, and responsible AI adoption.
原文の言語: 英語
Defines metrics, dashboards, reporting logic, and decision analytics.
原文の言語: 英語
Designs APIs, versioning strategy, request/response contracts, and integration patterns.
原文の言語: 英語
Coordinates architecture decisions, governance, and change across programs.
原文の言語: 英語
Reviews architectures for fitness, tradeoffs, anti-patterns, and delivery readiness.
原文の言語: 英語
Designs automation workflows, trigger logic, and task orchestration.
原文の言語: 英語
Designs backend systems, APIs, services, data access, and scalability patterns.
原文の言語: 英語
Acts as a business analyst who clarifies requirements, processes, and functional needs.
原文の言語: 英語
Maps capabilities, maturity levels, gaps, and enablement priorities.
原文の言語: 英語
Plans stakeholder adoption, communication, training, and behavior change.
原文の言語: 英語
Designs cloud architecture, landing zones, resilience, and platform choices.
原文の言語: 英語
Builds community programs, engagement loops, and member value systems.
原文の言語: 英語
Advises on compliance, controls, auditability, and regulated operating models.
原文の言語: 英語
Produces content systems, drafts, and channel-specific content outputs.
原文の言語: 英語
Designs content strategy, editorial direction, and content production systems.
原文の言語: 英語