timbrinded

This skill should be used when the user asks to "clean up", "remove slop", "simplify", "unslop", "reduce complexity", "tighten up", "cleanup before PR", "review for unnecessary complexity", "too verbose", "over-engineered", "remove unnecessary code", "AI generated mess", "simplify this branch", "reduce abstractions", "test bloat", "bloated tests", "test case bloat", or tests that only verify library or dependency behavior. Also use it for a dedicated simplification pass after AI generation. Do not use it as a substitute for correctness, security, performance, or architecture review. Report findings by default and apply fixes only when explicitly requested.

This skill should be used when the user asks to "create a Lottie", "generate a Lottie animation", "make a Bodymovin JSON", "fix a Lottie", "make a seamless loop", "make a large Lottie", "make a 30 MB Lottie", or mentions lottie-web, Bodymovin, Skottie, dotLottie, .lottie, or Lottie JSON.

This skill should be used when the user asks to create an infographic from repository documentation, Markdown docs, ADRs, specs, diagrams, research summaries, runbooks, or inline context. It writes copy-paste-ready image prompts by default, and if the user explicitly asks to generate, render, make, or create the actual image and an image generation tool is available, it should pass the generated prompt to that tool.

Teach languages, frameworks, concepts, debugging approaches, and problem-solving in a tutoring style that prioritizes hints, guided questions, and mental models over direct implementation. Use when the user says things like "teach me", "help me learn", "walk me through", "give me hints", "explain this", "don't just solve it", "don't edit my code", or when the user is working through an exercise, tutorial, kata, unfamiliar codebase, or new framework and wants explanation-first help without unsupervised code changes.

A design review and UI improvement skill for frontend interfaces. This skill should be used when the user asks to "review a design", "review a screenshot", "improve the UI", "make this look better", "fix the design", "check spacing", "improve visual hierarchy", "choose colors", "pick fonts", "review my CSS", "review my Tailwind", "make this more professional", "design feedback", "fix the layout", "improve typography", "check contrast", "add depth", "polish the UI", or is working on frontend code (HTML, CSS, JSX, TSX, Svelte, Vue) where design quality matters.

This skill should be used when the user asks to "optimize", "make faster", "speed up", "reduce memory", "improve performance", "reduce latency", "improve throughput", "profile", "benchmark", or mentions "slow code", "too slow", "performance bottleneck", "high memory usage", "allocation pressure", "GC pressure", "cache misses", or "hot path". Also applicable when reviewing code for performance characteristics, estimating the cost of an operation, or selecting data structures for performance.

This skill should be used when the user asks to "review SwiftUI code", "check HIG compliance", "audit iOS design quality", "generate SwiftUI views", "create an iOS screen", "improve SwiftUI design", "iterate on SwiftUI feedback", "fix accessibility in SwiftUI", "modernize SwiftUI code", "check deprecated APIs", "review this screenshot for iOS design", "add animations to SwiftUI", "add haptics to SwiftUI", or is working with SwiftUI views where design quality, accessibility, or Apple HIG compliance matters. Provides 40 enforceable rules, design token references, modern API guidance, animation/haptics patterns, and visual review rubrics for iOS design excellence.

This skill should be used only when explicitly asked to create or refresh a whole-repository authored Shape model from a fixed repository baseline. Survey architecture-significant subsystems, treat generated AST as read-only navigation, verify claims in source, and author validated model batches. Do not use it for incremental edits, pending-change review, or preflight planning.

This skill should be used before editing code to map a concrete planned change onto an existing Shape model, identify constraints, obligations, source anchors, coverage, and bindings, and optionally simulate a conservative temporary contract change. Do not use it for post-change review, whole-repository modeling, or source-to-model drift.

This skill should be used to review a concrete code diff for real reachable bugs, combining complete local review with focused authored-Shape investigation. Do not use it for pure model drift, authored contract-diff risk, style preferences, or speculative architecture concerns.

This skill should be used to compare a base and candidate authored Shape contract for semantic loosening that may remain checker-valid. Review `.shape` and vendored-pack changes only. Do not use it for application-code bugs, source-to-model drift, or whole-repository modeling.

This skill should be used for incremental Shape language work: authoring or repairing `.shape` files, explaining syntax or diagnostics, formatting, validation, teaching, source-to-model drift review, and CLI/editor operation. Do not use it for whole-repository modeling (`shape-index`), pre-implementation planning (`shape-contract-preflight`), authored contract-diff risk (`shape-contract-guard`), or code-change bug review (`shape-review`).

Phase 1 (preparation, not scored). On top of an already-generated whole-codebase AST Shape context, author higher-level architecture, boundary, and INVARIANT shapes that a downstream reviewer can use — grounded in the generated AST anchors. Built directly on the upstream shape-lang skill and the shp CLI.

Phase 2 (scored). Review a PR diff for real bugs. Exhaustive diff-only recall pass first (prune nothing), augment with cited cross-object bugs from a whole-codebase Shape model (mounted at ./shape, shp CLI on PATH), then a closed-list reality gate plus a human-salience prediction gate (emit the bugs a careful human would flag; never drop a severe bug). Promoted from variant v15-human-salience-gate: best generalized result at n=3 (+8.7 F1 overall vs same-model baseline; positive in all 5 repos; +10 on grafana/sentry). Emits review.json.

Act on selected code-review findings from pi-workflow-engine: make minimal fixes or post GitHub PR inline comments using gh, GitHub MCP/tools, or project-specific tools.

Operate and troubleshoot the hlgo Hyperliquid CLI for agent-native trading workflows. Trigger this skill for ANY task involving: hlgo command selection, flag lookup, JSON output or error contract parsing, dry-run planning, signing-path safety, debugging hlgo exit codes, composing multi-step trading workflows, integrating hlgo into pipelines, HIP-3 perp dex resolution, price/size precision questions, or generating scripts that call hlgo info/order/position/agent/account/config/version commands.

Comprehensive Parallel.ai Search API, Extract API, and Task API implementation and debugging guidance. Use when Codex needs to build, inspect, troubleshoot, or refactor integrations using parallel.ai / Parallel / parallel-web, Search API (/v1beta/search), Extract API (/v1beta/extract), Task API runs (/v1/tasks/runs), task groups/events (/v1beta/tasks/groups), SSE streams, webhooks, TaskSpec structured outputs, SourcePolicy filtering, processor selection, fetch policies, or Parallel beta headers.

Strongly opinionated Effect-TS coding, refactoring, debugging, and testing guidance for TypeScript projects using the `effect` ecosystem. Use when the user is working with Effect, Effect.Schema, `Context.Tag`, `Layer`, fibers/concurrency, `Stream`/`Sink`, `@effect/vitest`, `@effect/platform`, retries/schedules, or migrating from Promise/fp-ts/ZIO-style code to idiomatic Effect. Do not use for generic English uses of the word effect unrelated to the Effect-TS library.

This skill should be used when the user asks to "look up a Wikipedia article", "search Wikipedia for", "find the Wikipedia page on", "check Wikipedia about", or needs a quick fact or single article from the local Wikipedia data. For quick lookups and single-article reads — not for cross-domain exploration.