App Bug Sweep
Use when reviewing a built or shipped app (especially multi-platform — iOS + Android + web) for the class of latent defects that compile fine, pass unit tests, and only surface when a real user touches the UI — dead/no-op buttons, capability built on the backend but never wired to a screen, gestures declared but inert, silent data-loss footguns, and settings/data that are written-but-never-read o…
Open Multi Agent
Use when building autonomous agent systems in TypeScript/Node.js — triaging a list (GitHub issues, files, URLs) and dispatching parallel workers per item, running each worker inside its own git worktree or sandbox, fanning out one ephemeral agent per dynamic input via AgentPool, decomposing a goal into a parallel task DAG, running local gemma/qwen via Ollama or vLLM with concurrency caps, mixing…
One Item Per File
Use when writing or restructuring a non-trivial chunk of code (a module, a subsystem, an algorithm) and you want it precise, parallelizable, and maintainable — treat code as data. Enforces a three-phase authoring workflow — (1) define the data structures, (2) define the function signatures in terms of those structures so the skeleton compiles, (3) implement the functions — with strict one-item-pe…
Skills
Agent Skills published by @pjankiewicz, in the open Agent Skills format. Currently: open-multi-agent (TypeScript multi-agent orchestration via @jackchen_me/open-multi-agent).
Ux Review
Use when auditing or reviewing the UX/interaction quality of a mobile (iOS/Android) or web app — checking whether screens have distinct pressed/focus/disabled states, proper empty/loading/error states, accessible touch targets, platform-native conventions (Apple HIG / Material 3), WCAG 2.2 AA contrast, honest system status feedback, and clear copy. Provides a per-screen PASS/FAIL heuristic rubric…
Faithful Port
Use when porting or translating code from a cited reference into another language, framework, or engine (C++ → Rust, GLSL → WGSL, one renderer's/library's algorithm into yours) and faithfulness to the original is the goal. Builds on the one-item-per-file skill (data structures → signatures → implementations, one item per file) and adds reference-fidelity discipline: transcribe each reference func…