mit einem Klick
gon-stack
gon-stack enthält 28 gesammelte Skills von polubis, mit Repository-Berufsabdeckung und Skill-Detailseiten auf SkillsMP.
Skills in diesem Repository
Learning memos: concept → use case → example. Output markdown | mdx | html. Back = scenario OR code snippet. Trigger: memo, flashcard, /do-memo. Too broad → split first.
Conversion copywriter for homepage/landing/pricing/feature/about page copy. Use for "write copy," "improve this copy," "rewrite this page," "headline help," "CTA copy," "value prop," "tagline," "above the fold," "this copy is weak," "make this more compelling." Not for email copy (emails skill), popup copy (popups skill), pure line-edit of existing copy (copy-editing skill).
React UI component best practices for building consistent, accessible, and maintainable component libraries. Use this skill when writing, reviewing, or optimizing React components, styling patterns, or component APIs.
Accessibility auditor for HTML markup or live URLs. Trigger on "check accessibility", "audit a11y", "review HTML", "check-html", or pasted URL/code for a11y review.
Requirements Inference & Specification Architect. Use it when you're not sure that your requirements are clear enough for the engineering team to start working on them. This skill will ask you a series of questions to clarify any ambiguities in your requirements and produce a final, detailed specification document.
Generate a commitlint-compliant commit message with a description, prompt user for approval, and run git commit. Use when user asks to commit, create a commit, generate a commit message, or wants to stage and commit changes. NEVER runs git push. Uses git diff --cached to understand staged changes.
Generate a comprehensive UX/UI design document from the product MVP definition. Use when designing UI/UX, defining design tokens, creating view layouts, or when asked to create a ux-ui.md document.
Requirements Inference & Specification Architect. Use it when you're not sure that your requirements are clear enough for the engineering team to start working on them. This skill will ask you a series of questions to clarify any ambiguities in your requirements and produce a final, detailed specification document.
Use when Polish text needs language and typography corrections — fixes missing Polish characters, typography, and tech-term Polonizations. Asks user whether output should be plain text or markdown before returning result.
Implement a feature from requirements in the exact style of an ideal example. Trigger on "clone", "implement like", "use X as example", "copy this module's style", "clone <module> and implement <feature>", or "build feature matching <example>". Requires both requirements and an ideal-example reference.
Scan an ideal-example module and write an AGENTS.md describing its architecture rules and code conventions, with links and a hashy-stamped version. Trigger on "document module", "create AGENTS.md", "describe conventions", "scan module and list rules", "make ideal-example doc". This is the step-b generator for ideal examples.
Orchestrate a full feature through staged planning and parallel implementation by sequencing the other framework skills. Trigger on "feature workflow", "run the full workflow", "implement this epic end to end", "/feature-workflow", or when a large feature must go through requirements → plans → implementation → integration → review.
Stamp or verify a markdown doc's version/hash against the code it documents using the @repo/hashy CLI. Trigger on "hash this doc", "stamp AGENTS.md", "update doc hash", "hashy", "version this markdown", "check doc drift", or after editing an ideal-example doc.
Plan domain models from requirements — pure TypeScript types for domain entities and events, decoupled from backend and persistence shapes. Trigger on "plan domain", "domain models", "model the domain", "domain types", "design domain layer", or after backend/db planning to define the in-app domain.
Plan a frontend module's architecture from requirements — views, components, state/handlers, and API communication across presentation/core/integration layers. Trigger on "plan frontend", "plan FE", "frontend architecture", "plan the UI module", "design the views and state". Not for graphic→element analysis (use plan-ui) or graphic→code (use do-ui).
UI plan generator from screenshot or mockup. Trigger when user wants to analyse or document UI elements from a graphic. Graphic required; user description optional. Output flat element list with behaviors and variants — no code.
Pixel-perfect UI generator from graphic reference. Trigger when user asks to recreate UI from design/screenshot/mockup, or mentions "pixel perfect". Require graphic first. Output plain Tailwind UI with single component entry `<ComponentName />`.
Plan backend APIs from requirements. Ask API style (REST/RPC), suggest realtime (SSE/WS) when needed, output Zod `in`/`out` contracts for endpoints/procedures/channels.
Use when designing a database schema from requirements. Validates all required info upfront, then outputs complete database plan in one shot — entities, diagrams, indexes, security, normalization, migration files, schema doc. Trigger on "design schema", "database design", "db schema", "model the data", "schema for feature", "design the tables".
Plan feature requirements from user input into standardized markdown and auto-fill missing structural gaps from provided context. Trigger on prompts like "plan requirements", "standardize requirements", "make feature plan", "fill missing requirement sections", or "prepare final feature doc".
Make Technical Interview report. Takes raw interviewer notes about a candidate and produces a fully filled markdown report using the standard interview report template. Trigger when user provides interview notes, Q&A logs, or candidate feedback and wants a formatted report.
Turn requirements into Supabase DB design with strict approval gates (schema, RLS/access, RPC, indexes, jobs, migration plan) using concise output.
Build tests from user test plan using strict black-box behavior. Trigger when user asks to generate tests from requirements/spec and avoid implementation leaks. Report untestable requirements with reason, then map each requirement to test status table.
Use when creating/extending e2e tests with step interpreter and shared commands, where test engine + interpreter path/contract must be confirmed before execution.
Batch post language polisher. Reads a file (or inline content) with posts separated by "
Gathers and validates UI component requirements, asks clarifying questions for missing API/behavior/demo details, requests branding and code-style sources, proposes an implementation plan for approval, then builds pure UI components and optional demo pages in the requested location. Use when the user asks to craft, build, or generate UI components.
Orchestrate asynchronous task implementation from a TODO markdown file by launching subagents in separate git worktrees. Use when the user invokes async-implement, asks to execute TODO.md tasks in parallel, or requests multi-agent implementation with isolated branches/worktrees.
Postgres performance optimization and best practices from Supabase. Use this skill when writing, reviewing, or optimizing Postgres queries, schema designs, or database configurations.