hjemmesidekongen
114 verified skills
bitbucket-basics
Bitbucket PR creation, pipeline status checks, and code review actions via the Atlassian MCP server. Use when creating pull requests, checking CI/CD pipeline status, adding PR comments, approving PRs, listing open PRs, or performing branch operations against a Bitbucket Cloud or Data Center workspace.
tools
file-ownership
Decompose a feature into parallel development streams with strict file boundary ownership. Each stream owns specific files exclusively; interface contracts are defined at boundaries before implementation begins. Prevents merge conflicts structurally. Use when decomposing a feature for parallel agent execution, when planning a wave where multiple agents will work simultaneously, when a previous parallel attempt produced merge conflicts, or when a multi-stream plan needs explicit ownership assignments to prevent write contention.
development
mcp-creator
Create and configure MCP server integrations for Claude Code plugins. Use when adding external service connections, setting up .mcp.json, configuring stdio/SSE/HTTP/WebSocket servers, integrating MCP tools into commands and agents, debugging MCP server connectivity issues, adding tool annotations, or configuring OAuth and API key authentication for MCP servers.
tools
prompt-optimizer
Sharpens vague prompts into precise instructions using proven frameworks. Mode A (auto-sharpen): hook-triggered, infers missing context from codebase, rewrites the prompt, shows the sharpened version, proceeds. Mode B (prompt builder): command-triggered via /prompt:create, takes rough intent, applies best framework, outputs a polished prompt for copy or execution.
development
verification-gate
Mandatory 5-step verification protocol before claiming any task is done. IDENTIFY the proof command → RUN it → READ the full output → VERIFY it matches expected → ONLY THEN claim done. Bans inference-based completion ("should work", "probably", "seems to", "likely", "I believe it works"). Use when completing any task in a plan wave, finishing a skill or hook implementation, before marking a wave done, or before reporting success on any multi-step change.
testing
content-writer
Drafts brand-aware written content: READMEs, landing pages, blog posts, social posts, product descriptions, email newsletters, and marketing copy. Applies brand voice, SEO, and copywriting best practices with automated quality checks. Use for any content writing or copywriting task.
testing
api-security
Rate limiting, input validation, CORS, security headers, API keys, and request sanitization.
development
expo-api-routes
API routes in Expo Router — server functions, request handling, middleware, and type-safe endpoints
development
expo-deployment
EAS Build, OTA updates, app store submission, and Expo deployment workflows.
development
http-protocols
HTTP/1.1 vs 2 vs 3, TCP handshake, cookies, sessions, CORS, caching headers, and protocol fundamentals. Covers multiplexing, QUIC, Set-Cookie attributes, JWT vs server sessions, preflight mechanics, Cache-Control directives, and content negotiation.
development
nuxt-patterns
Nuxt 3/4 patterns — auto-imports, useFetch, server routes, layers, and module development
development
package-managers
pnpm workspaces, npm scripts, lockfile management, peer dependencies, and node version management. Covers package manager selection, monorepo setup, lockfile discipline, peer dep resolution, and version pinning strategy.
testing
skill-quality-rubric
8-dimension quality framework for evaluating smedjen knowledge skills. Provides a scoring template for audits with pass/fail thresholds and actionable remediation guidance.
development
tdd-workflow
Red-green-refactor cycle enforcement for test-driven development. Guides writing failing tests first, making them pass with minimal code, then refactoring.
development
reasoning-trace-optimizer
Analyzes agent reasoning patterns in trace-light.log to detect context degradation, tool confusion, instruction drift, goal abandonment, circular reasoning, and premature conclusions. Produces a diagnosis with specific patterns found and a fix recommendation. Use when an agent session is underperforming, when repeated tool calls aren't making progress, when an agent seems to have lost track of its goal, when debugging why a plan stalled or looped, or when reviewing a completed session for quality issues.
tools
openapi-typescript
OpenAPI code generation, type-safe API clients, schema validation, and contract-first development. Covers openapi-typescript, openapi-fetch, zod-openapi, breaking change detection, and CI integration.
tools
expo-cicd
CI/CD with EAS, GitHub Actions, app distribution, and automated build/deploy pipelines for Expo.
development
expo-data-fetching
Data fetching in Expo — fetch, React Query, SWR, offline support, and caching strategies
development
expo-dev-client
Custom Expo dev client setup, native module debugging, and development build workflows for projects that need native code beyond Expo Go's scope.
tools
expo-jetpack-compose
Jetpack Compose integration with Expo modules — native Android views, bridging patterns, and module API
development
expo-native-ui
Native UI patterns in Expo — platform-specific components, gestures, animations, and design system integration.
testing
expo-swiftui
SwiftUI integration with Expo modules — native iOS views, bridging patterns, and module API
development
expo-tailwind
NativeWind / Tailwind CSS in React Native — setup, styling patterns, theme config, and platform differences for Expo projects.
development
nextjs-app-router
Next.js App Router patterns — server components, route handlers, caching, layouts, and data fetching.
data-ai
session-handoff-herold
Save active task context for cross-session continuity — captures task state, subtask progress, blockers, and open questions. Extends kronen session-handoff with herold-specific fields. Detects prior handoffs on resume.
testing
brainstorm-session
Run an open-ended brainstorm session where Claude acts as a sparring partner. Challenges assumptions, pushes back on weak reasoning, and follows the user's lead. No predefined questions or structured output. Use when exploring a problem space, thinking through architectural decisions, comparing tradeoffs between approaches, challenging assumptions about a design, or hashing out options before committing to a direction.
testing
design-loader
Loads existing design artifacts from .ai/design/{name}/ and reports what's available. Checks for tokens.yml, identity.yml, platform tokens, and contrast matrix. Use when checking design status, loading design context for downstream skills, or running /segl:status.
testing
hook-reviewer
Review and validate Claude Code hooks for correctness, security, and reliability. Use when reviewing hook scripts after creation, auditing existing hooks for issues, validating hook changes before committing, checking exit codes and JSON output schemas, diagnosing silent or failing hooks, or verifying hook security against injection and path traversal.
development
jira-basics
Basic Jira operations via Atlassian MCP — read tickets, search with JQL, transition status, add comments. Use when asked to check a Jira ticket, search issues, update ticket status, or add a comment in Jira.
tools
parallel-reviewer
Dispatch 4 specialized reviewers in parallel (security, performance, architecture, testing), each analyzing the code independently. Coordinator consolidates findings, deduplicates, and ranks by severity. Use when reviewing a non-trivial code change that touches multiple quality dimensions, before merging a feature branch, when a PR needs thorough multi-angle review, or when a single reviewer pass may miss domain-specific issues.
development
root-cause-debugging
4-phase mandatory investigation protocol before any fix is attempted: Investigate → Pattern Analysis → Hypothesis → Implement. Prevents band-aid fixes by requiring evidence gathering, pattern identification, and a testable hypothesis before touching code. Includes escalation tree, anti-patterns (shotgun debugging, fix-and-pray, symptom fixation), symptom-to-tool mapping, and framework-aware debugging references. If 3+ fixes fail, escalate to architecture review. Use when diagnosing a bug or unexpected behavior, when a quick fix isn't working and the problem keeps returning, when a test is failing and the cause isn't immediately obvious, when a system behaves differently than expected, or when debugging any framework-specific issue (load the relevant tech skill's references/debugging.md for framework detail).
tools
session-handoff
Create and resume from handoff documents for cross-session context continuity. Use when: saving state before ending a session, context window approaching capacity, major milestone completed, resuming previous work ("continue where we left off"), or after 5+ file edits in a session. Complements snapshot.yml with human-readable chained handoffs and staleness awareness.
testing
slack-basics
Slack channel reading, message search, and posting via Slack MCP. Use when reading channel history, searching for past messages, posting updates to a channel, replying to threads, listing channels, or looking up users by name or email.
tools
pencil-tokens
Bridges design tokens to Pencil. Reads tokens.yml and guideline.yml, resolves all YAML references to concrete values, maps semantic colors + typography + spacing + radius + shadows to Pencil's set_variables format, then calls set_variables. Light mode only. One-shot token injection — run before any Pencil design work.
development
seo-fundamentals
SEO fundamentals — E-E-A-T quality signals, Core Web Vitals, technical SEO, on-page optimization, keyword strategy, meta tag rules, structured data, and the SEO Health Index scoring model (0-100 weighted composite).
development
social-media-patterns
Social media content patterns — platform-specific strategies for LinkedIn, Twitter/X, Instagram, TikTok. Post templates, hook formulas, content pillars, repurposing workflows, and engagement optimization.
testing
auth-patterns
OAuth 2.0, JWT, sessions, RBAC, refresh tokens, PKCE, and authentication architecture.
tools
e2e-testing-patterns
End-to-end testing patterns with Playwright and Cypress. Page object model, test fixtures, assertions, waiting strategies, and CI integration.
testing
expo-dom-components
Web components in native Expo context — DOM component bridges, WebView patterns, and hybrid rendering
development
expo-simulators
iOS Simulator and Android Emulator management — setup, screenshots, recording, and debugging tools.
tools
expo-upgrading
Expo SDK version migration — upgrade workflow, breaking changes, native module compatibility, and rollback strategy.
development
git-advanced-workflows
Advanced git workflows: branching strategies, interactive rebase, bisect, cherry-pick, stashing, worktrees, and conflict resolution patterns.
tools
nodejs-patterns
Node.js streams, worker threads, event loop, error handling, and clustering patterns. Covers when to use streams vs buffers, CPU-bound offloading decisions, async error propagation, and graceful shutdown.
testing
react-patterns
React hooks, context, composition, performance optimization, and common anti-patterns. Covers when to use each hook, composition over inheritance, performance decision tree, and state management choices.
development
project-mapper
Scan a repository to detect tech stack, workspace shape, and produce a project profile at .ai/project-map.yml. Includes skills matching: maps detected stack to relevant smedjen knowledge skills for reference_paths dispatch. Use when onboarding to a codebase, before agent dispatch, when checking which skills apply, or generating a project profile.
development
task-decomposer
Decomposes task descriptions into ordered subtask lists with complexity estimates, file scope, dependency chains, and 5-factor risk scores. Handles ambiguity by requesting clarification when task scope is insufficiently defined.
testing
typescript-modern
Modern TypeScript — utility types, generics, discriminated unions, template literals, satisfies, and type-level programming. Covers narrowing strategy, generics vs overloads, utility type selection, and type safety rules.
tools
vue-patterns
Vue 3 Composition API, reactivity system, Pinia state management, and Vue 2→3 migration patterns.
development
brand-strategy
Creates a brand from scratch through market research and interactive sparring. Runs competitive research via Perplexity, then guides the user through positioning, audience, voice, values, and content pillars. Produces the full brand guideline set at .ai/brand/{name}/. Use when building a new brand, defining brand strategy for a product, or when /våbenskjold:create is invoked.
development
command-creator
Create, modify, or debug Claude Code slash commands with correct frontmatter, dynamic arguments, file references, and bash execution. Use when adding commands to any plugin, fixing broken or invisible commands, configuring argument-hint or positional arguments, namespacing commands with subdirectory prefixes, or building multi-step command workflows.
tools
hypothesis-generator
Generate three competing hypotheses for a debugging problem, then investigate each in parallel and synthesize evidence to identify the root cause. Extends root-cause-debugging with structured hypothesis management. Use when a bug's root cause is unclear after initial evidence gathering, when a previous fix failed and the cause is still unknown, when multiple subsystems could plausibly be at fault, or when confirmation bias in investigation needs to be avoided.
development
instinct-evolve
Review qualified instincts and propose evolution into skills, rules, or memory entries. Reads .ai/instincts/instincts.yml and promotes high-confidence patterns to durable artifacts. Human approves each promotion. Use when: promoting instincts to skills, promoting patterns to CLAUDE.md rules, running the evolution phase of the learning pipeline, reviewing which instincts qualify for evolution, or closing out the instinct review cycle.
devops
brand-audit
Codifies an existing brand from materials, samples, and references. Analyzes provided content to extract voice patterns, values, and positioning. Produces the same guideline format as brand-strategy. Use when a brand already exists but isn't documented, or when /våbenskjold:audit is invoked.
development
brand-evolve
Guided reinvention of an existing brand guideline. Loads current brand from .ai/brand/{name}/, identifies what to keep vs change, and walks the user through targeted evolution. Preserves brand equity while updating positioning, voice, or values. Use when refreshing a brand or when /våbenskjold:evolve is invoked.
documentation
brand-loader
Loads brand guidelines from .ai/brand/{name}/ and makes them available to the current context. Progressive disclosure: L1 confirms brand exists, L2 loads summary, L3 loads specific files on demand. Use when a downstream skill or user needs brand context, or when /våbenskjold:apply is invoked.
testing
accessibility-wcag
WCAG 2.1 AA compliance — ARIA, keyboard navigation, screen readers, color contrast, and semantic HTML. Covers the POUR principles, semantic-first authoring, ARIA as last resort, and practical testing workflows.
development
agent-creator
Create autonomous Claude Code agents with correct frontmatter, triggering examples, system prompts, and tool restrictions. Use when adding agents to any plugin, defining agent behavior, writing system prompts, configuring model tier and tool access, setting up auto-triggering descriptions, or choosing between agent vs command for a task.
tools
agent-dispatcher
Dispatches decomposed subtasks to worker agents with tier-appropriate model assignment. Enforces file ownership, includes reference_paths from project profile for knowledge loading, collects reports, retries failures once at the same tier, then escalates to the next tier. Called after task-decomposer and tier-assignment complete.
data-ai
agent-reviewer
Review and validate Claude Code agents for frontmatter correctness, description quality, system prompt structure, tool scope appropriateness, and model tier justification. Use when reviewing agents after creation, auditing existing agents for compliance, validating agent changes before committing, checking triggering examples, or auditing all agents in a plugin for consistency.
tools
agent-teams
Parallel team execution patterns for multi-agent workflows. Use when planning parallel agent dispatch, choosing team compositions (review, debug, feature, research, security teams), or integrating with Agent Teams experimental API. Covers file ownership rules, task coordination, and integration patterns.
development
auto-doc
Automatically update documentation after task completion. Reads git diff to identify what changed, then updates CLAUDE.md structure tree and counts, ecosystem.json entries, MEMORY.md facts, and roadmap.yml statuses. Use when doc-checkpoint reports stale docs, after multi-file tasks, after adding or removing components, or when documentation has drifted from reality.
testing
azure-devops-pipeline
Query Azure DevOps pipeline status, trigger runs, and parse test results for the active task branch. Requires the 'azure-devops' MCP server. Falls back to a clear not-configured message when unavailable.
tools
bitbucket-pr-workflow
Create PRs from the active task with structured description auto-populated from the task key, summary, acceptance criteria, and test plan. Checks pipeline status after creation and links back to Jira with a comment. Falls back gracefully when Bitbucket or Jira MCP is unavailable.
tools
brainstorm-decision-reader
Load and filter past brainstorm decisions relevant to the current task. Scans all decisions.yml files across brainstorm topics, filters by domain relevance, and returns a structured summary. Use when starting new work, before architectural decisions, when any skill or agent needs prior context about what was already decided, checking what decisions exist for a domain, or reviewing past conclusions before making new ones.
testing
brainstorm-decision-writer
Extracts structured decisions from a brainstorm session. Walks through the conversation with the user, identifies landing points, and co-authors a decisions.yml file with domain tags and confidence levels. Use when formalizing brainstorm conclusions, extracting decisions from transcripts, recording what was decided after a discussion, detecting contradictions with previous decisions, or closing out an active brainstorm session.
testing
brand-voice-implementation
Brand voice implementation in code — formality spectrum, tone attributes, personality archetypes, UI copy tone selection (plain speech vs brand voice), channel-specific guidelines, and voice consistency checking across components.
development
bulk-ingestion
Batch ingest tickets from a Jira board or filter URL — fetches all assigned non-closed tickets with contradiction detection on each. Requires the Atlassian MCP server. Falls back gracefully when unavailable.
tools
c4-architecture
Generate C4 model architecture documentation with Mermaid diagrams at four levels: system context, container, component, and deployment. Use when documenting how systems interact at different abstraction levels, creating stakeholder-appropriate architecture views, or mapping deployment topology. For general-purpose diagrams (flowcharts, sequence, ERD, class), use mermaid-diagrams instead.
devops
code-review
Bidirectional code review — supports requesting reviews from the code-reviewer agent and processing received review feedback. Categorizes findings by severity (critical, important, minor) with actionable, direct language.
development
completion-gate
10-point quality gate run by the code-reviewer agent after implementation. Verifies spec compliance, lint, tests, security, and visual output. The implementing agent never runs this gate on its own work.
development
confluence-basics
Confluence page search and content reading via Atlassian MCP. Use when asked to find documentation, check Confluence, search by title or content, read page content, navigate page hierarchies, or filter by space or label.
tools
hook-creator
Create, modify, or debug Claude Code hooks with correct event types, exit codes, JSON output formats, and shell best practices. Use when adding hooks to any plugin, fixing broken or silent hooks, blocking or allowing specific tools via PreToolUse, adding guards for SessionStart or Stop events, registering hooks in plugin.json, or converting hooks between types (command, prompt, HTTP, agent).
tools
verification-before-completion
Mandatory verification protocol before claiming any task is done. Run tests, check output, verify edge cases, and confirm acceptance criteria are met.
testing
confluence-lookup
Search Confluence for pages related to the active task. Extracts keywords from task summary, description, and labels. Constructs CQL queries, scores results by relevance, and links matching docs to the task context for dev-engine consumption. Graceful fallback when Confluence MCP is not configured.
tools
content-strategy-patterns
Content strategy patterns — pillar strategy (40/25/25/10), editorial calendars, content templates (blog, social, email, video), repurposing matrix, and pre-publishing checklists. Planning framework for consistent content production.
development
contradiction-detection
Analyze Jira ticket descriptions against comments to detect contradictions, modifications, and extensions to requirements. Produces a severity-ranked list of findings with original text, contradicting text, and resolution suggestions. Runs standalone or auto-triggered after jira-ingestion.
tools
design-to-code-patterns
Translating design specs into frontend components — design token consumption, spacing/typography/color systems, responsive breakpoints, component decomposition from mockups, and design-code handoff conventions.
development
design-tokens
Transforms visual identity tokens.yml into platform-consumable formats: Tailwind theme JSON, CSS custom properties, and DTCG JSON. Generates WCAG contrast matrix with colorblind safety notes. Use when generating platform tokens from visual identity, creating Tailwind theme, building CSS variables, or running /segl:tokens.
development
doc-checkpoint
Evaluate documentation state after task completion. Compares changed files against CLAUDE.md, MEMORY.md, roadmap.yml, and ecosystem.json to produce a structured checklist of needed updates. Use after multi-file tasks, roadmap deliveries, when adding or removing skills, commands, hooks, or agents, checking if docs are stale, or verifying ecosystem.json registration matches files on disk.
testing
error-handling-patterns
Consistent error handling across codebases. Error boundaries, try-catch patterns, custom error types, logging strategies, and graceful degradation.
development
eslint-config
ESLint flat config, custom rules, plugin setup, overrides, and migration from legacy .eslintrc format. Covers rule selection, TypeScript integration, Prettier compatibility, and CI performance.
tools
fumadocs-patterns
Fumadocs documentation framework — MDX components, content structure, meta.json navigation, LLM integration (llms.txt, getLLMText), and docs site conventions. Use when building or editing documentation pages, configuring Fumadocs components, or setting up AI-readable doc endpoints.
development
git-worktree-isolation
Structured git worktree lifecycle for isolated branch work. Creates a second working tree at .worktrees/<name>/ on a new branch, auto-detects and runs setup, verifies clean baseline. Finishes with 4 options: merge to main, open PR, keep branch, or discard. Safety: always verifies .worktrees/ is gitignored before creating. Use when starting risky or experimental work that shouldn't affect the main working tree, when parallel feature branches need isolation, before large refactors that may need to be discarded, or when a clean baseline is needed to test something without stashing.
development
marketing-psychology-patterns
Marketing psychology and strategy patterns — PLFS scoring for behavioral models, MFS scoring for marketing ideas, journey-stage model selection, ethical guardrails, and stage-based prioritization (pre-launch through scale).
data-ai
instinct-extractor
Extract atomic instincts from raw interaction observations. Reads .ai/instincts/observations.jsonl, identifies recurring patterns, and writes structured instincts to .ai/instincts/instincts.yml. Updates confidence scores and marks contradictions. Use when: reviewing accumulated observations, running the learning pipeline, updating instinct confidence, extracting patterns from recent session data, or seeding new instincts before instinct-evolve.
devops
jira-ingestion
Fetch Jira tickets via Atlassian MCP and normalize to local YAML task files. Single ticket mode (by key) or bulk mode (by JQL filter). Extracts key, summary, description, acceptance criteria, comments, attachments, status, and assignee. Auto-runs contradiction-detection after ingestion. Falls back to dry-run mode with sample data when MCP is unavailable.
tools
mermaid-diagrams
Create software diagrams using Mermaid syntax. Use when generating class diagrams, sequence diagrams, flowcharts, ERDs, state diagrams, git graphs, or gantt charts. For C4 architecture diagrams specifically, use c4-architecture instead.
testing
nestjs-patterns
NestJS modules, providers, guards, interceptors, pipes, and testing patterns. Module organization, request lifecycle, decorator patterns, and test strategy.
testing
outlook-basics
Email search, thread reading, and calendar access via Microsoft 365 MCP. Use when searching for emails by sender, subject, or keyword, reading an email thread, checking a calendar for upcoming meetings, listing meetings in a time range, or finding free/busy slots.
tools
package-audit
npm audit workflow, outdated dependency triage, license compliance checking, and update strategies for patch, minor, and major version bumps. Covers Renovate/Dependabot config and monorepo update coordination.
testing
plan-engine
Goal-driven iterative planning engine. OODA loop that plans one wave at a time, executes via multi-agent dispatch, and accumulates learnings across cycles. Always iterative, always goal-driven. Use when breaking work into waves, planning multi-file changes, or running structured execution.
development
playwright-testing
Playwright locators, assertions, fixtures, visual comparison, trace viewer, and CI setup.
testing
plugin-creator
Create Claude Code plugins with correct directory structure, plugin.json manifest, component organization, and marketplace registration. Use when scaffolding new plugins, adding components to existing plugins, troubleshooting plugin loading or registration issues, configuring ecosystem.json, publishing to the marketplace, or setting up plugin.json hooks and metadata.
tools
plugin-reviewer
Review and validate Claude Code plugins for manifest correctness, component registration, portability, and settings patterns. Use when reviewing plugins after creation, auditing existing plugins, validating plugin changes before committing, checking ecosystem.json against files on disk, verifying marketplace registration, or auditing for hardcoded paths and portability issues.
tools
prisma-patterns
Prisma schema design, migrations, relations, raw queries, seeding, and performance optimization. Covers schema modeling decisions, migration workflow, query patterns, transactions, and common anti-patterns.
testing
project-config
Project-level configuration via .ai/project.yml — profile presets (work/personal), hook flag overrides, dev commands, and git conventions. Explains the unified config system: how to bind a profile, which flags are overridable vs safety-locked, and where to find the full schema. Use when setting up a new project, switching profiles, configuring hook behavior, checking which profile is active, understanding override precedence, or asking about project.yml fields.
development
project-profile-loader
Load project-level configuration from .ai/project.yml. Provides project name, context (work/personal), profile presets, git conventions, QA format, dev commands, and MCP server list. Determines which completion gate checks are relevant for the active project. Loaded at session start or task switch.
tools
qa-handover-generator
Generate structured QA handover notes for completed tasks. Content includes what changed, test scenarios, screenshots (if UI), and regression risks. Output format determined by project profile: Jira comment, GitHub PR description, or plain markdown. Triggered by task completion.
development
reducing-entropy
Delete dead code, remove unused files, and shrink the codebase. Applies a deletion-first mindset: measure success by how much code remains, not effort spent. Use when cleaning up after a feature, removing unused skills or plugins, simplifying over-engineered code, auditing for dead imports or unreachable paths, or deciding whether a refactor actually reduces complexity. Manual activation only.
tools
roadmap-capture
Automatically captures out-of-scope ideas, deferred features, and future work items into the project roadmap during conversation. Use when an idea surfaces that is outside the current task scope and should be tracked for later, when the user says "add this to the roadmap" or "track this for later", or when a deferred item from a brainstorm or plan needs persistent tracking.
data-ai
security-audit
OWASP Top 10 vulnerability assessment, dependency scanning, secret detection, CSP and security header review, SAST/DAST guidance, and supply chain security. Use when hardening a codebase before release, reviewing a PR with auth or input handling changes, or responding to a reported vulnerability.
development
sitemap-planning
Information architecture and sitemap planning — page hierarchy, content siloing, header structure, internal linking strategy, schema markup selection, URL conventions, and featured snippet optimization.
development
skill-creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, optimize a skill's description for better triggering accuracy, or run a TDD baseline eval before writing a skill to target real behavior violations.
testing
skill-reviewer
Review and validate Claude Code skills for structure, content quality, and progressive disclosure. Use when reviewing skills after creation, auditing existing skills for issues, validating skill changes before committing, checking description trigger quality, verifying frontmatter compliance, or auditing all skills in a plugin for consistency.
tools
storybook-patterns
Storybook stories, args, decorators, interaction tests, visual testing, and addon configuration. Use when writing CSF3 stories, setting up play functions for interaction tests, configuring addons, integrating with Chromatic for visual regression, writing MDX documentation, or structuring stories for React, Vue, or Angular.
development
tailwind-v4
Tailwind CSS v4 — migration from v3, CSS-first theme configuration, custom utilities, dark mode strategies, container queries, and responsive design. Covers @theme, @utility, @variant, @apply usage, and PostCSS integration.
tools
tier-assignment
Evaluates decomposed subtasks using a weighted 5-factor risk matrix and assigns model tiers (junior/senior/principal). Produces per-subtask justifications to support dispatch decisions and cost optimization.
testing
turborepo-patterns
Turborepo pipeline config, caching, filtering, remote cache, and monorepo task orchestration. Use when configuring turbo.json pipelines, setting up remote cache, filtering tasks to specific packages, migrating from Lerna or Nx, pruning for Docker builds, or debugging cache misses.
development
web-copywriting
Web copywriting patterns — headlines, CTAs, microcopy, error messages, empty states, and landing page structure. Covers above-fold framework, benefit-driven copy, the Seven Sweeps editing method, and UI text guidelines.
development
visual-identity
Creates a complete visual identity system from brand direction or user input. Generates color palettes (OKLCH-based, 10-stop scales), typography pairing, spacing system, and foundational design decisions. Outputs tokens.yml as the primary design truth file. Use when creating visual identity from brand guidelines, building color palettes, selecting typography, or running /segl:identity.
development
visual-verification
Captures Playwright screenshots at 3 breakpoints and verifies UI output against design intent. Three verification tiers: automated layout checks, LLM vision comparison, and human escalation for low-confidence results.
testing
vite-patterns
Vite configuration, plugins, HMR, build optimization, and environment variables.
tools
writing-clearly-and-concisely
Write clear, concise prose for humans — documentation, commit messages, error messages, reports, UI text. Applies Strunk's rules and avoids AI writing patterns. Use when writing or editing any text that humans will read.
testing
web-security-owasp
XSS, CSRF, injection attacks, SSRF, security headers, and web application security hardening. Covers reflected/stored/DOM-based XSS, token-based CSRF defense, parameterized queries, SSRF allowlists, and the full security header stack with recommended values.
development