kalilurrahman/kr-claudiator-skills - Page 6

launch-playbook

Plan a tiered launch (alpha → GA) with success metrics and rollback gates.

llm-eval-harness

Build an offline LLM eval harness with golden sets, judges, and regression gates.

llm-structured-output

Reliably extract structured data from LLM responses using function calling, JSON mode, and validation. Outputs schema definitions, extraction patterns, retry logic, and parsing pipelines.

long-context-1m-tokens

Design prompts that exploit Claude 5 Mythos/Fable's 1M+ token window for whole-codebase reasoning, multi-doc synthesis, and cross-file refactors without lossy summarisation.

mcp-tool-orchestration

Wire Claude 5 to MCP servers (filesystem, git, DB, Jira, Notion) with typed tool contracts, per-tool guardrails, error taxonomies, and traceable retries.

metric-tree-builder

Decompose a north-star metric into input/output trees with owners and alert thresholds.

microservices-communication

Design communication patterns between microservices. Outputs synchronous vs asynchronous decision framework, service discovery approach, circuit breaker configuration, and observability strategy.

ml-feature-selection

Select the most predictive features for ML models to improve accuracy and reduce overfitting. Outputs filter, wrapper, and embedded selection methods with validation strategy.

ml-feature-store

Stand up a feature store with offline/online parity and point-in-time correctness.

mobile-security

Implement mobile application security for iOS and Android. Outputs secure storage, certificate pinning, code obfuscation, biometric auth, and API security patterns.

model-cards

Author a model card covering intended use, evals, risks, and update cadence.

model-compression

Compress and optimise ML models for production deployment. Outputs quantisation, pruning, and distillation approaches with size-accuracy tradeoff analysis.

model-fine-tuning-strategy

Plan and execute model fine-tuning to improve performance on domain-specific tasks. Outputs fine-tuning decision framework, dataset preparation guidelines, evaluation strategy, and deployment plan.

monorepo-strategy

Design a polyglot monorepo with Turbo/Nx, shared packages, hermetic builds, and selective CI.

multi-region-failover

Architect multi-region active-active or active-passive with traffic steering.

multimodal-doc-extraction

Extract structured data from mixed-media documents (scanned invoices, forms, ID cards, spreadsheets embedded in PDFs) with per-field confidence and human-review routing.

null-object-pattern

Eliminate null checks and NullPointerExceptions using the Null Object pattern. Outputs null object implementations, optional type usage, and refactoring strategies for null-heavy codebases.

oauth-security

Secure OAuth 2.0 implementations against common attacks. Outputs threat model, security controls, token security hardening, and implementation review checklist.

observer-pattern

Implement the Observer pattern for event-driven decoupled communication. Outputs event bus implementations, typed event systems, and async observer patterns.

okr-cascade

Translate company objectives into team-level OKRs with measurable key results.

openapi-design

Design and document REST APIs using the OpenAPI 3.1 specification — schemas, security definitions, response codes, examples, and SDK generation. Outputs complete OpenAPI specs, documentation, and validation tooling.

operating-rhythm

Design a quarterly operating rhythm: planning, reviews, retros, and metrics.

opportunity-solution-tree

Build and use Opportunity Solution Trees to connect product outcomes to customer opportunities and solutions. Outputs tree structure, prioritisation process, assumption mapping, and experiment design.

opportunity-tree

Build an opportunity-solution tree from a desired outcome and discovery insights.

pdf-diagram-reasoning

Reason over PDFs, architecture diagrams, flowcharts, and whiteboard photos — extracting entities, relationships, and decisions with page-grounded citations.

phishing-simulation

Run a phishing simulation program with cohorts, lures, and learning paths.

platform-idp

Design an Internal Developer Platform (IDP) that reduces cognitive load for engineers. Outputs platform capabilities, golden paths, self-service workflows, and adoption strategy.

platform-strategy

Develop a platform strategy that creates an ecosystem around your product. Outputs platform definition, ecosystem design, API strategy, marketplace model, and network effect cultivation.

platform-team-charter

Charter a platform team: product mode, paved roads, and golden paths.

pricing-experiment

Design a pricing experiment with willingness-to-pay tests and guardrails.

product-analytics

Build product analytics infrastructure to understand user behaviour, feature adoption, and business outcomes. Outputs event taxonomy, funnel analysis, retention queries, and reporting infrastructure.

product-health-metrics

Define a comprehensive product health dashboard tracking acquisition, engagement, retention, and quality. Outputs metric definitions, threshold setting, anomaly detection, and weekly review process.

product-led-growth

Design and implement a product-led growth strategy where the product drives user acquisition, expansion, and retention. Outputs PLG model, activation metrics, viral loops, and freemium conversion strategy.

product-teardown

Conduct a structured product teardown analyzing UX, positioning, business model, growth strategy, and technical decisions. Outputs a framework-grounded analysis with strategic observations and lessons applicable to your own product.

progressive-delivery

Implement progressive delivery strategies including canary releases, feature flags, and traffic splitting. Outputs deployment pipeline, rollout configuration, automated rollback triggers, and observability requirements.

prompt-injection-defense

Defend LLM-powered applications against prompt injection attacks. Outputs threat model, input/output sanitisation, privilege separation, detection patterns, and monitoring strategy.

prompt-versioning

Manage prompt versions with diffs, evals, and traffic shifts across environments.

query-cost-governance

Apply warehouse cost controls: slot caps, scan budgets, query tags, and chargeback.

queue-vs-stream-decision

Choose between queues and streams given ordering, retention, and replay needs.

raci-builder

Build a tight RACI matrix for a cross-team initiative.