All authors
Claude Skills by emanuelrechsteiner
github.com/emanuelrechsteiner54 skills0 installs4 views
- Daily DocsDaily logbook entry — item counting runs exclusively through a reference script (never manually/by feel), aggregates signals/git/memory for the prose, writes markdown + syncs to Notion.Votes: 0GitHub stars: 2
- Nightly ObservationNightly housekeeping — calls the safe rotate-signals.sh (archive-then-trim + pre-truncate backup + count-conservation abort), then compute-daily-metrics.sh for the metrics row, health check. Re-enabled 2026-06-21 after IMP-049 fix.Votes: 0GitHub stars: 2
- Weekly ImproveWeekly pattern extraction — reads 7 days of signals, extracts recurring patterns, and writes IMP proposals to ~/.claude/plans/ for human review.Votes: 0GitHub stars: 2
- Audit ConfigRun a deterministic audit of ~/.claude configuration against known schema-drift patterns (invalid settings fields, stale model IDs, broken hook patterns, naming collisions). Triggers on "audit config", "check config", "verify configuration", "config audit", "konfiguration prüfen", "audit my claude setup", "config check", or quarterly via launchd. Reports CRITICAL/WARN findings to ~/.claude/audit-reports/.Votes: 0GitHub stars: 2
- Backend DevelopmentFirebase and state management specialist. Handles Firestore operations, Firebase Functions, Zustand stores, user-scoped security, and optimistic updates with rollback. Use when working with Firebase, Firestore, Cloud Functions, Zustand, state management, APIs, or backend logic. Triggers on "Firebase", "Firestore", "Zustand", "state management", "backend", "API", "Cloud Functions", "security rules", "database", "store", "backend bauen", "datenbank anbinden", "firestore regeln", "cloud function...Votes: 0GitHub stars: 2
- Check ParallelizableAnalyze whether a task can be parallelized into independent sub-agents. Returns a structured proposal with file claim-sets, agent assignments, and confirmation prompt — or a clear explanation why sequential is better. Auto-triggers BEFORE any non-trivial task that involves multiple files or units. Triggers on 'check parallel', 'analyze parallel', 'kann ich das parallel', 'can this be parallel', 'parallelization analysis', 'geht das parallel', 'parallel prüfen', 'parallelisierbar', 'lässt sich...Votes: 0GitHub stars: 2
- Cloud Cli DisciplineGuidance for inspecting cloud-CLI state before destructive or scope-changing operations (env rm, link --yes, --force) and avoiding team/account auto-pick; covers when --yes is safe, preferring the web UI for per-environment values, and the Vercel team-mismatch pre-deploy check (vercel whoami, teams ls, env ls, switch). Use before running a destructive or scope-changing cloud-CLI command, or when a deploy might target the wrong team or account. Triggers on "vercel", "vercel deploy", "vercel li...Votes: 0GitHub stars: 2
- Create HookCreate Claude Code hooks. Use when the user wants to create a hook, register a hook in settings.json, add lifecycle automation (SessionStart, PreToolUse, PostToolUse, Stop), gate or audit tool calls, or block specific commands. Asks for missing requirements, then creates the hook script + registration.Votes: 0GitHub stars: 2
- Create RuleCreate Claude Code rules for persistent agent guidance. Use when adding coding standards, project conventions, security policies, workflow rules, or any guidance you want auto-loaded every session. Asks for missing scope/purpose, then writes a focused rule file to ~/.claude/rules/ (user-global) or .claude/rules/ (project).Votes: 0GitHub stars: 2
- Create SkillCreate Claude Code Skills (SKILL.md files). Use when authoring a new skill, when the user asks about SKILL.md structure or skill frontmatter, when packaging reusable workflows or domain knowledge into a forked-context capability. Skills differ from rules (always-loaded) and hooks (event-triggered) by being on-demand.Votes: 0GitHub stars: 2
- Create SubagentCreate custom Claude Code subagents (specialized AI agents with focused system prompts and tool allowlists). Use when adding a new agent type like code-reviewer, debugger, doc-writer, or domain-specific assistant. Asks for missing scope/purpose, then writes the agent .md file with proper frontmatter.Votes: 0GitHub stars: 2
- Dependency AuditAudit npm dependencies for security vulnerabilities, outdated packages, and unused dependencies. Provides actionable upgrade recommendations. Use to maintain healthy dependencies. Triggers on "dependency audit", "npm audit", "outdated packages", "security vulnerabilities", "dependency check", "update dependencies", "dependencies prüfen", "abhängigkeiten aktualisieren", "veraltete pakete", "sicherheitslücken prüfen", "npm pakete checken", "ungenutzte abhängigkeiten".Votes: 0GitHub stars: 2
- DocumentationTechnical documentation specialist. Maintains API docs, architecture documentation, developer guides, JSDoc, README updates, and Storybook stories. Use when creating or updating documentation, writing guides, documenting APIs, or maintaining project documentation. Triggers on "document", "documentation", "README", "API docs", "developer guide", "JSDoc", "architecture docs", "write docs".Votes: 0GitHub stars: 2
- Find DocsRetrieves authoritative, up-to-date technical documentation, API references, configuration details, and code examples for any developer technology. Use this skill whenever answering technical questions or writing code that interacts with external technologies. This includes libraries, frameworks, programming languages, SDKs, APIs, CLI tools, cloud services, infrastructure tools, and developer platforms. Common scenarios: - looking up API endpoints, classes, functions, or method parameters -...Votes: 0GitHub stars: 2
- Fix ReviewPost-fix completeness checker. After making a code fix, this skill identifies the fix scope, searches for all related occurrences, verifies call sites are updated, and checks for cascading issues. Use after fixing bugs to ensure nothing was missed. Triggers on "review fix", "check fix", "fix complete", "verify fix", "did I miss anything", "cascading changes", "fix überprüfen", "fix checken", "nichts vergessen", "noch mehr stellen", "habe ich was übersehen", "gibt es noch weitere", "alle stell...Votes: 0GitHub stars: 2
- Framework ExtractionFramework-in-Project pattern — develop a new framework (rules, commands, skills) INSIDE a host project first, then extract only when the 3-criteria trigger fires (stability + reuse need + pain frequency). Use when building or evolving a framework inside a host project, or deciding whether/how to extract it to a standalone repo. Triggers on "framework extraction", "extract framework", "framework-in-project", "when to extract", "extraction trigger", "extraction checklist", "standalone repo", "f...Votes: 0GitHub stars: 2
- GrillingInterview the user relentlessly about a plan, decision, or idea until every branch of the decision tree is resolved — one question at a time, each led with a concrete recommendation. Use before large or ambiguous undertakings, when requirements feel underspecified, or on 'grill me', 'stress-test my plan', 'löcher mich', 'hinterfrag meinen plan', 'denk das mit mir durch', 'frag mich aus'.Votes: 0GitHub stars: 2
- Historical Signals V2Enriched historical signal extraction with full tool-use coverage (Bash, Read, Edit, etc.), tool-results, and SQLite storage. Use for deep retrospective analyses requiring relational queries (co-edit graphs, tool-use sequences, error patterns). Triggers on 'deep historical analysis', 'tool-use sequences', 'co-edit graph', 'error pattern extraction', 'cross-source signal join'.Votes: 0GitHub stars: 2
- Historical SignalsConvert historical Claude Code JSONL session files into signals.jsonl format for meta-observer ingestion. Use when running historical observation analysis, ingesting past session data into the improvement pipeline, performing multi-month/year deep-dive aggregations, or comparing cross-project tool-use trends. Triggers on 'historical analysis', 'analyze past sessions', 'cross-project aggregation', 'multi-year signal trends', 'ingest session history'.Votes: 0GitHub stars: 2
- Human TestingSenior UX/UI tester that performs comprehensive manual testing using Playwright MCP. Navigates through the entire app, clicks all buttons, tests all functions, takes screenshots, checks visual consistency, validates styling principles, and reports issues. Use when you need thorough app testing, visual regression checks, or UX validation. Triggers on "human test", "manual test", "test the app", "QA test", "visual test", "click through", "test all functions", "UX test", "UI test", "teste die ap...Votes: 0GitHub stars: 2
- I18n Bilingual SyncMirror locale JSON edits across a bilingual pair (e.g. de.json ↔ en.json) so adding/modifying/removing a key in one file is reflected in the sibling. Triggers on "translate this key", "sync locale", "i18n sync", "add translation", "de.json", "en.json", "übersetzung ergänzen", "übersetze diesen key", "locale synchronisieren", "übersetzungen abgleichen", "sprachdateien synchron halten", "fehlende übersetzungen", or when editing one locale file in a pair.Votes: 0GitHub stars: 2
- Import FixerFix broken imports after file moves or renames. Scans for import errors, identifies broken paths, and updates all affected files. Use after moving files, renaming modules, or when seeing "Cannot find module" errors. Triggers on "fix imports", "broken imports", "cannot find module", "update imports", "import error", "imports reparieren", "kaputte imports", "modul nicht gefunden", "imports fixen", "importpfade aktualisieren", "import fehler".Votes: 0GitHub stars: 2
- Incident ResponseIncident-response runbook for production issues: observe via read-only diagnostics, form one falsifiable hypothesis, verify it against source before touching anything, ship a minimal hotfix, write a mandatory regression test, then postmortem. Diagnosis stays strictly read-only until the hypothesis is confirmed in code — jumping to a fix on a guess is the core failure mode this skill prevents. Production mutations are always ESCALATE-band, even mid-incident. Backs /launch-team's Run mode; also...Votes: 0GitHub stars: 2
- Ios Simulator TestingDocuments the iOS Simulator toolchain (MCP simulator control + build tools, xcrun simctl, xcodebuild) for building, launching, and interacting with native iOS apps in the Simulator across any project on this Mac. Use when the user wants to run, test, or debug an app on the iOS Simulator. Triggers on "iOS simulator", "run on iPhone", "test on simulator", "simulator testen", "app im simulator", "nativ testen", "iphone testen", "simulator screenshot", "xcodebuild", "simctl".Votes: 0GitHub stars: 2
- Kokonutui ProKokonutUI Pro setup and component installation — per-project shadcn/ui registry config, KOKO_PRO_TOKEN handling, and the complete component slug index (references/index.md). Use when setting up KokonutUI Pro in a project, adding components, or debugging registry auth/404 errors. Triggers on "kokonutui", "@kokonutui-pro", "koko", "KOKO_PRO_TOKEN", "shadcn add @kokonutui-pro", "kokonutui component", "kokonutui einrichten", "kokonutui komponente hinzufügen", "koko pro registry".Votes: 0GitHub stars: 2
- Legacy Codebase AuditOnboarding audit checklist for TypeScript/React codebases flagged as legacy, built with less powerful agents, or showing quality-debt signals. Use on the first session in an unfamiliar TS/React codebase or before feature work on quality-debt code. Triggers on "audit", "codebase audit", "legacy", "technical debt", "modernization", "built with older/less powerful agents", "quality debt", "altlast", "altlasten", "technische schulden", "modernisierung", "code-audit", "codebasis prüfen", "aufräume...Votes: 0GitHub stars: 2
- Memory IndexCross-project memory query layer. Aggregates the 35+ project-specific memory directories under ~/.claude/projects/ into on-demand MCP Memory Graph entities. Enables queries like "what did I learn about Next.js middleware?" or "which projects use SwiftData?". Triggers on "memory search", "find project context", "cross-project query", "what have I learned about X", "search projects", "remember Y".Votes: 0GitHub stars: 2
- Meta ObserverOn-demand meta analysis of observation signals. Reads signals.jsonl, session-metrics.jsonl, improvement-ledger.json, and git logs to synthesize reusable patterns and IMP-XXX improvement candidates. Replaces the archived improvement-agent Meta Layer. Triggers on "meta observe", "analyze observations", "extract improvements", "synthesize patterns", "what have I learned".Votes: 0GitHub stars: 2
- Migrate To SkillsMigrate legacy rule/command files (Cursor .mdc rules, slash-command markdown, old-format skills) to Claude Code SKILL.md format. Use when consolidating cross-platform configs, importing Cursor or Antigravity workflows to Claude Code, or refactoring old rules into the skills/ directory.Votes: 0GitHub stars: 2
- Navigate ModeCodebase navigation primitives for fast orientation. Bundles canned scripts to locate entry points, list routes, list state stores, and surface recently-touched hot files. Use instead of ad-hoc grep when answering "where is X" / "how is this wired" questions. Triggers on "where is", "find entry point", "list routes", "list stores", "show me the routes", "hot files", "what's the main file", "navigate", "navigation", "orient me", "wo ist", "einstiegspunkt", "wo liegt".Votes: 0GitHub stars: 2
- Nextjs DebugFramework diagnostics for Next.js applications. Troubleshoots API route 404/405 errors, middleware redirect loops, SSR/hydration issues, and server/client component problems. Use when facing Next.js-specific issues. Triggers on "nextjs debug", "404 error", "405 error", "middleware issue", "hydration error", "server component", "client component", "API route not working".Votes: 0GitHub stars: 2
- OrchestrationScrum master and orchestrator for multi-agent coordination. Manages agent workflows, enforces quality gates, coordinates handoffs, and ensures proper sequencing. Use when coordinating multiple specialized tasks, managing complex features, or when work spans backend, frontend, and testing. Triggers on "orchestrate", "coordinate", "multi-agent", "workflow", "manage agents", "complex feature", "scrum", "sprint", "quality gate".Votes: 0GitHub stars: 2
- Parallel DispatchOrchestrate N parallel write-mode sub-agents safely. Validates disjoint file sets, pre-claims locks per agent, dispatches parallel Agent calls in ONE message, synthesizes results. Use when you need true parallelism with write access — e.g. refactor 5 routes simultaneously, add tests to 4 different modules in parallel, migrate N components at once. Triggers on 'parallel dispatch', 'parallel refactor', 'run agents in parallel', 'fan out work', 'parallel ausführen', 'mehrere agents gleichzeitig'...Votes: 0GitHub stars: 2
- Pattern DocumentLearning extraction skill. Analyzes recent fixes, extracts reusable patterns or anti-patterns, and creates rule documents in ~/.claude/rules/. Use after discovering bugs or implementing solutions that could help prevent future issues. Triggers on "document pattern", "extract learning", "create rule", "prevent this bug", "pattern document", "save this pattern", "was haben wir daraus gelernt", "muster dokumentieren", "regel erstellen", "daraus eine regel machen", "dieses problem nie wieder", "s...Votes: 0GitHub stars: 2
- Project BootstrapRepository exploration and project bootstrap specialist. Explores repos, audits docs, and optionally bootstraps Memory MCP context. Use when starting a new project, exploring a codebase, or setting up project structure. Triggers on "bootstrap", "explore repo", "new project", "codebase exploration", "project setup", "audit docs", "repo erkunden", "projekt aufsetzen", "codebase erkunden", "neues projekt starten", "projektstruktur anlegen", "verschaff dir einen überblick".Votes: 0GitHub stars: 2
- Project PlanningStrategic project architect using Claude Opus for comprehensive planning. Creates multi-agent development plans, coordinates task distribution, and ensures agent collaboration with zero conflicts. Use when planning new features, creating project architecture, or designing implementation strategies. Triggers on "plan", "planning", "architecture", "strategy", "roadmap", "feature design", "implementation plan", "project structure".Votes: 0GitHub stars: 2
- PrototypeBuild a throwaway prototype to answer a design question — a runnable interactive terminal app for state/logic questions, or several radically different UI variations on one route switchable via URL param. Use when unsure whether a state model or logic feels right, or what a UI should look like. Triggers on 'prototype', 'spike', 'wegwerf-prototyp', 'varianten bauen', 'mal durchspielen', 'ausprobieren bevor wir bauen', 'welche variante'.Votes: 0GitHub stars: 2
- Quality ReviewMilestone-level multi-specialist code review. Dispatches 6 parallel specialist agents (architecture, security, performance, testing, maintainability, docs) on a targeted file set and synthesizes findings by severity. Use when a feature/milestone feels complete and you want a thorough multi-angle quality pass before commit/PR. Triggers on 'quality review', 'review my work', 'deep review', 'milestone review', 'qualitäts review', 'gründlich prüfen'.Votes: 0GitHub stars: 2
- Rcode IosiOS-specific adaptations of the R.Code atomic development workflow (toolchain defaults, SwiftData patterns, Swift Testing, mandatory i18n gate, /phase-gate overrides). Use when working in a R.Code-managed iOS project — `.rcode/` exists alongside `.xcodeproj` or `Package.swift`. Triggers on "iOS", "SwiftData", "SwiftUI", "Swift Testing", "SwiftLint", ".xcodeproj", "Package.swift", "Xcode", "@Observable", "Localizable.xcstrings", "iOS app bauen", "iOS projekt", "iOS entwickeln", "swift tests sc...Votes: 0GitHub stars: 2
- Rcode OnboardThis skill should be used when the user asks to 'onboard', 'get started', 'what should I work on', 'pick up project', 'where are we', 'continue project', 'what is the current state', 'wo stehen wir', or 'was ist der stand'. It reads project status, architecture, conventions, and recent agent activity to get a new agent productive in under 10 minutes on a R.Code-managed project.Votes: 0GitHub stars: 2
- React Perf CheckPerformance anti-pattern detection for React components. Detects state/ref conflicts, unnecessary re-renders, missing memoization, and animation performance issues. Use when experiencing UI lag, jank, or unexpected behavior. Triggers on "react performance", "slow render", "re-render", "laggy", "jank", "animation slow", "state ref conflict", "useEffect performance", "react performance prüfen", "langsames rendern", "ui ruckelt", "unnötige re-renders", "animation ruckelt", "komponente ist langsam".Votes: 0GitHub stars: 2
- Release Cli DisciplineRelease and CLI value-handling discipline — local-first deploy testing, tarball-based publish verification, safe printf-based CLI/secret piping, and MCP generation-endpoint serialization. Use before deploying a serverless function/build, publishing a package, writing env vars/secrets into a CLI, or invoking a generation-style MCP tool. Triggers on "deploy", "vercel deploy", "build locally", "serverless function", "npm publish", "pnpm pack", "npm pack", "changeset publish", "tarball", "release...Votes: 0GitHub stars: 2
- ResearchDocumentation research specialist using web fetching and search. Researches technologies, APIs, best practices, and gathers implementation guides. Use when learning about new technologies, finding documentation, researching solutions, or gathering best practices. Triggers on "research", "documentation", "learn about", "how to", "best practices", "API docs", "Firebase docs", "React docs", "find documentation", "investigate", "recherchiere", "recherche", "nachschlagen", "erkläre mir wie", "wie ...Votes: 0GitHub stars: 2
- Resolving Merge ConflictsWork through an in-progress git merge or rebase conflict hunk by hunk, resolving by original intent traced to each side's primary sources (commits, PRs, issues), then run the project's checks and finish the operation. Use on merge/rebase conflicts, 'resolve conflict', 'merge-konflikt', 'rebase-konflikt', 'konflikt auflösen', 'CONFLICT (content)'.Votes: 0GitHub stars: 2
- Scope CheckThis skill should be used when the user asks to 'check scope', 'scope check', 'verify scope', 'scope review', or mentions 'scope creep', 'scope shrinkage', 'out of scope', 'scope prüfen', 'scope kontrolle', 'bleiben wir im scope', 'zu viel gemacht', 'haben wir uns verzettelt', 'while we are here', 'mission creep', 'feature creep'. It verifies work matches the original plan by detecting unplanned additions (scope creep) and dropped features (scope shrinkage). Works at per-unit and per-phase gr...Votes: 0GitHub stars: 2
- Scroll Animation PatternsReusable scroll-linked animation patterns for React/Next.js. Covers RAF-driven character-chaos effects, sticky card decks, scroll-snap overlays, and avoids common state/ref conflicts. Use when implementing scroll-linked effects, hero reveal animations, or text/character staggered animations. Triggers on "scroll animation", "sticky card", "parallax", "character chaos", "scroll-linked", "RAF animation", "hero animation".Votes: 0GitHub stars: 2
- Tailwindcss V4 StylingTailwindCSS v4 specialist ensuring correct theme configuration, variable naming, and utility class usage. Validates @theme directives, prevents spacing/container naming conflicts, and enforces v4 best practices. Use when setting up TailwindCSS v4, creating theme customizations, debugging styling issues, or implementing UI components with Tailwind. Triggers on "TailwindCSS", "Tailwind v4", "@theme", "styling broken", "CSS not working", "max-w", "spacing utilities", "theme configuration", "styl...Votes: 0GitHub stars: 2
- Testing SuiteQuality assurance specialist ensuring 90%+ test coverage. Creates unit tests with Vitest, E2E tests with Playwright, and visual tests with Storybook. Use when writing tests, validating code quality, checking test coverage, or ensuring accessibility compliance. Triggers on "test", "testing", "unit test", "E2E", "coverage", "Vitest", "Playwright", "Storybook", "QA", "quality assurance", "accessibility", "WCAG", "tests schreiben", "testabdeckung", "coverage checken", "testsuite", "unit tests anl...Votes: 0GitHub stars: 2
- Type CoverageTrack and improve TypeScript type coverage. Identifies untyped code, implicit any usage, and type safety gaps. Use to understand codebase type safety and systematically improve it. Triggers on "type coverage", "typescript coverage", "any types", "type safety", "improve types", "type audit".Votes: 0GitHub stars: 2
- Ui DevelopmentReact component implementation specialist. Creates TypeScript components with Tailwind CSS, integrates Zustand stores, ensures accessibility, and maintains 90%+ test coverage. Use when building React components, implementing UI, working with Tailwind, or creating frontend interfaces. Triggers on "React", "component", "UI", "Tailwind", "frontend", "interface", "form", "button", "modal", "Storybook".Votes: 0GitHub stars: 2