AI Agent Workflow Collection

Writes in ASD-STE100 simplified technical English — short sentences, plain words, active voice. Use when the user wants terse technical prose, names this skill, or asks for STE100.

Disables the AI's default hedging, diplomatic softening, and social-nicety layers to deliver cold, unfiltered, consequentialist analysis on any topic. Use when the user wants the AI's actual assessment without padding, regardless of how harsh or uncomfortable the honest answer is.

Applies a modern CSS reset covering box-sizing, margin removal, text rendering, media defaults, and root stacking context. Use when starting a new web project, creating a global stylesheet, or when the codebase lacks sensible CSS defaults.

Applies depth-driven layered color and shadow principles to elevate average UI into polished, readable interfaces for both light and dark themes. Use when improving or reviewing UI components, setting up a design token system, applying consistent depth and shadow to an interface, or making a flat or inconsistent UI feel more refined without a full redesign.

Applies Chesterton's Fence, Occam's Razor, the Peter Principle, and Dunning-Kruger to technical decisions. Use before refactoring, architectural changes, or any task where changing existing behavior carries risk or requires judgment about what the user actually needs.

Stops "just one more" small tweaks that add scope without changing the outcome. Use when named (just-stop), or when about to implement extra polish, cleanup, or drive-by fixes beyond the stated goal.

Autonomous build-verify-fix loop that runs until an explicit Definition of Done passes. Use when named (mini-loop) or when the user says exactly "run a loop".

Run a 3-leg evaluation against a newly released LLM. Leg 1 is trap questions with deterministic answers, leg 2 is an agentic orchestration task with a planted failure, leg 3 is a brownfield feature build in a real repo. Use when the user wants to vibe-check, test, evaluate, or compare a new frontier model, or mentions model-bench, model vibe check, or launch-day protocol.

Prisma ORM patterns including Prisma Client usage, queries, mutations, relations, transactions, and schema management. Use when working with Prisma database operations or schema definitions.

Stress-tests ideas by surfacing assumptions, counterarguments, and confidence gaps. Use when evaluating claims, proposals, plans, or decisions that need rigorous intellectual scrutiny.

Use when creating UI components in React - functional components, hooks, custom hooks, or component composition patterns. NOT when backend logic, API routes, or non-React frameworks are involved. Triggers: "create component", "build widget", "KPI card", "form", "modal", "custom hook", "useContext", "useState", "useEffect".

Applies Parkinson's Law, the Pareto Principle, and Goodhart's Law to produce tight, high-impact responses. Use when you want the agent to cut padding, lead with what matters, and avoid optimizing for appearance over substance.

Cuts fluff and leads with what matters. Applies Parkinson's Law, Pareto, and Goodhart's Law so every word earns its place. Use when the user wants terse output, names this skill, or signals that responses are too padded.

Generates as many digits of pi as possible and writes them into files inside a tokenMaxxing/ folder. Use when the user asks to token-maxx, fill a file with pi digits, test agent output throughput, or generate N pi files sequentially or in parallel.

Runs a comprehensive 20-point code quality audit on a "vibe-coded" codebase that was shipped fast and now needs hardening. Use this skill whenever the user mentions vibe coding, vibe audit, code audit, code review, code cleanup, hardening a prototype, productionizing a codebase, technical debt cleanup, or asks Claude to "audit my code", "review my codebase", "clean up my project", or check a freshly-shipped project for common quality issues. Also use when a user shares a repo and asks what's wrong with it or how to make it production-ready.

Vermeidet KI-typische Schreibmuster und erzeugt hochwertige Sach- und Fachtexte, die gelesen, gespeichert und erinnert werden. Trigger bei Anfragen wie "schreibe einen Artikel", "erstelle einen LinkedIn-Post", "verfasse einen Blog-Beitrag", "schreibe einen Text ueber", "ueberarbeite diesen Artikel", "schreibe einen Fachtext", "erstelle einen Sachtext", "schreibe etwas ueber". Auch triggern bei "mach den Text menschlicher", "klingt zu sehr nach KI", "natuerlicher schreiben". Basiert auf Wikipedia Signs of AI Writing (August 2025) und Journalismus-Forschung zu Leser-Retention.