kayaman

Guidance for designing charts, graphs, plots, dashboards, and data visualizations that communicate clearly and persuade. Use when creating or reviewing a visualization, choosing a chart type, picking a color palette, decluttering a busy graphic, fixing misleading axes or proportions, building a dashboard, annotating a figure, or turning data into a presentation, report, or data-driven story. Grounded in the standard data-visualization literature (Knaflic, Tufte, Cleveland & McGill, Cairo, Wilke, Munzner, Few, Berinato). Covers chart selection, graphical perception and encoding, color and accessibility, decluttering, graphical integrity, dashboards, and narrative. Does NOT cover building data pipelines or ETL, statistical modeling or analysis methods, BI tool/vendor selection, or general UI/UX layout (see ux-design-principles). Tool-agnostic, with optional Python recipes.

Architect and implement production-grade microservices systems in TypeScript (NestJS) and Python (FastAPI), including resilience, observability, testing, deployment, and migration guidance.

Design, configure, curate, govern, monitor, and integrate Databricks AI/BI Genie Spaces — the natural-language-to-SQL surface over Unity Catalog. Covers space scoping, general instructions, parameterized example SQL, SQL functions, trusted assets, JOIN configuration, knowledge store, certified queries, benchmarks, monitoring tab, feedback loops, the Genie Conversation API, governance via Unity Catalog (row filters, column masks, embedded warehouse credentials), and sharing/permissions. Use whenever the user mentions Genie, Genie Space, AI/BI Genie, Databricks natural language analytics, text-to-SQL on Databricks, or asks to build/curate/troubleshoot/embed a Genie experience. Do NOT use for: AI/BI Dashboards (visual BI without natural language), Databricks Assistant (in-IDE coding companion), Mosaic AI Agent Framework (custom agent apps), generic text-to-SQL outside Databricks, or LLM RAG over unstructured documents (Genie operates on structured Unity Catalog data only).

Implement OTP and passwordless authentication on AWS for TypeScript projects using Cognito CUSTOM_AUTH triggers (default) or a custom DynamoDB-backed flow, with SES (email) and SNS (SMS) delivery. Use when the user mentions OTP, one-time password, passwordless login, magic link, Cognito custom auth, DefineAuthChallenge, CreateAuthChallenge, VerifyAuthChallengeResponse, SES verification email, SNS SMS code, or MFA over email/SMS. Covers architecture decision (Cognito vs custom), Lambda trigger handlers, SES/SNS notifiers, DynamoDB schema with TTL, rate limiting, constant-time comparison, threat model (enumeration, replay, brute force), and aws-sdk-client-mock testing.

Apply UX psychology and usability principles when building user interfaces. Distills Laws of UX (Fitts's, Hick's, Jakob's, Miller's, Peak-End, Von Restorff, Tesler's, Doherty, Postel's, Aesthetic-Usability), Don't Make Me Think usability heuristics, and lightweight UX process methods into actionable code-level guidance. Use when building UI components, reviewing interfaces for usability, reducing cognitive load, designing navigation or forms, choosing interaction patterns, or auditing for dark patterns. Does NOT cover typography, color palettes, responsive CSS, or design tokens — see related skills.

Design, implement, and test event-driven systems — event modelling, schema evolution, transactional outbox, sagas (orchestration and choreography), event sourcing, CQRS, idempotent consumers, and tests that actually catch race conditions. Use whenever the user mentions events, event bus, message broker, Kafka, RabbitMQ, SNS/SQS, EventBridge, Pub/Sub, Kinesis, event sourcing, CQRS, sagas, choreography, outbox, CDC, dead-letter queue, eventual consistency, at-least-once delivery, or when designing asynchronous coupling between services. Also use when reviewing event-driven code for failure modes or converting a synchronous design to event-driven. Do NOT use for fire-and-forget in-process callbacks, UI event listeners, DOM events, plain pub/sub inside a single process, reactive UI state machines, or pure request/response API design (use api-design-principles instead). Do NOT use to pick a specific broker (Kafka vs RabbitMQ vs SQS) — this skill is vendor-neutral.

O'Reilly book reference lookup for software design decisions. Coding agents MUST use this skill whenever making or reviewing any design decision — choosing an architecture pattern, selecting a data structure, structuring a module, evaluating a library, deciding on an API contract, applying a design pattern, weighing trade-offs between approaches, or any moment where a choice between two or more implementation strategies comes up. The O'Reilly MCP handles the actual book search; this skill tells you how and when to invoke it. Trigger even for seemingly small decisions (naming, layering, concurrency model, error handling strategy) — the best engineers reach for authoritative references before committing to an approach.

Architecture and design-review skill covering Domain-Driven Design (DDD), Clean Architecture, SOLID, GoF design patterns, and event-driven architecture, with concrete Python 3.12+ and TypeScript 5.x idioms. Use whenever a structural design decision is in play — designing a new system or module, reviewing code for structural quality, choosing between architectural approaches, refactoring toward cleaner layering, placing business logic, defining module boundaries, evaluating aggregates and bounded contexts, applying SOLID, selecting a design pattern, or deciding when (and when not) to adopt event sourcing, CQRS, or event-driven messaging. Trigger on phrases like "design a system", "architecture review", "is this clean", "where does this logic belong", "aggregate", "bounded context", "hexagonal", "ports and adapters", "SOLID violation", "anemic domain model", "code smell", "refactor this", "use case layer", "event sourcing", "CQRS", or any moment a developer asks whether a piece of code is in the right place.

Draft or update an Architecture Decision Record (ADR) in docs/adrs/. Use whenever the user is choosing between design alternatives, introducing a new library/framework/database, defining a service boundary, reversing a past decision, or when a PR review surfaces "why did we do it this way?". Produces a numbered ADR file following the Nygard-style template (Context, Decision, Status, Consequences) plus an Alternatives Considered section.

Scan the codebase, PR, or recent conversation for the 20 individual-contributor and team-level anti-patterns catalogued in The Effective Software Engineer (Ch. 5 & Ch. 10). Use when the user asks "what smells?", "code smells", "am I doing anything wrong?", before a retrospective, when onboarding to a new codebase, or when a PR feels off. Produces a report listing each detected anti-pattern with evidence (file:line, commit, or quoted conversation) and the remedy from the book. Deliberately conservative — reports only anti-patterns it can cite evidence for.

Append a one-line accomplishment to docs/BRAG.md, or summarize entries from a time window. Use when the user says "log this", "add to my brag doc", "record this", after shipping something meaningful, or when prepping for a 1:1, self-review, or promo packet. Subcommands via first argument; `add` is default. Also invoked automatically by the Stop hook at the end of substantive turns (see .claude/hooks/brag-log.sh).

Stage and write a disciplined Conventional-Commit message for the current changes. Use when the user says "commit", "let's commit this", "make the commit", or when finishing a TDD loop. Inspects the staged and unstaged diffs, groups changes by concern, and either produces ONE commit if the diff is cohesive, or proposes a split into multiple commits if it is not. Never commits without showing the message first.

Record, list, or repay an item in the technical-debt ledger at docs/TECH_DEBT.md. Use whenever a shortcut is taken in the current change (hard-coded value, bypassed validation, skipped test, out-of-band config, copied-and-tweaked code), when the user says "let's come back to this" or "add a TODO", or when exploring the codebase reveals a debt item worth recording. Subcommands, selected via the first argument: `add`, `list`, `repay`, `audit`.

Drive a disciplined, hypothesis-based debugging loop for a bug, failing test, flaky test, or unexpected behavior. Use when the user reports "it's broken", pastes an error or stack trace, shows a failing test, or says "why is this happening?". Forces a written hypothesis before any code change, records what was tried, and converges by bisection rather than by guessing. Not for feature work — use /tdd for new behavior.

Reframe a task around its outcome before any code is written. Use at the start of any new feature, bug-fix, refactor, or when the user's request reads like an output ("add a button", "write a script", "build an endpoint") without a stated "so that…". Also use when the user asks "what should I work on?" or when scope feels fuzzy. The goal is a one-sentence outcome that names the user, the action, and the measurable result.

Produce the three artifacts that should exist AFTER code ships but are routinely skipped — a README/changelog update, an observability plan (metrics, logs, alerts), and a communication note (announcement draft for Slack/email/changelog). Use when a feature has been merged or is about to deploy, when the user says "we shipped it" or "announce this", or on your own initiative right after a /commit that closes a notable outcome. Explicitly refuses to treat the code-merge as "done".

Enforces Semantic Versioning 2.0.0 (semver.org) rules. Use when choosing a version bump type for a release, validating version strings, tagging releases, handling pre-release or build metadata, managing deprecations, or advising on version precedence.