Results for “spa”

Design-token systems — semantic CSS custom properties for colours, spacing, typography, and shadows, with light/dark theme switching via `[data-theme]` overrides. Use when setting up a token catalogue, auditing components for hardcoded colours/magic numbers, or adding dark-mode support via token swaps.

Drive the user's desktop in the background — clicking, typing, scrolling, dragging — without stealing the cursor, keyboard focus, or switching virtual desktops / Spaces. Cross-platform: macOS, Windows, Linux. Works with any tool-capable model. Load this skill whenever the `computer_use` tool is available.

Expert guidance for Ibis, the Python dataframe library that provides a pandas-like API but generates SQL for execution on any backend — DuckDB, PostgreSQL, BigQuery, Snowflake, Spark, and more. Helps developers write analytics code once and run it anywhere without rewriting SQL for each database.

Design, implement, and validate reproducible machine-learning pipelines spanning data preparation, training, evaluation, registry, and deployment gates. Use when the user requests an ML pipeline, needs to turn model scripts into an orchestrated workflow, or provides pipeline components that must be connected safely.

Orchestrate a fleet of parallel `codex exec` workers with you (Claude Code) as the supervisor — spawn one per isolated git worktree, dispatch headless, verify each INDEPENDENTLY, PR/merge. The manual "codex-ultracode" pattern for fanning out real implementation, research, or review work onto Codex. Bakes in the hard gotchas (stdin blocking, background tracking, don't-trust-self-reports, writer isolation). Triggers on — orchestrate codex, codex workers, codex fleet, spawn codex, delegate to codex in parallel, manual ultracode, 开 codex 小弟, 派 codex worker — NOT for a single cross-vendor opinion (use the `second-opinion-codex` agent), NOT for web-UI worker decomposition (use `/orchestrate`).

Create uncontested market space using value innovation instead of competing head-to-head. Covers the Four Actions Framework, Six Paths, strategy canvas, and buyer utility map.

Evaluates text-to-image generation models by decomposing assessment into five specialized skills (object presence, count, spatial relations, scale, and text rendering) and open-ended prompts, producing interpretable binary scores with visual and textual explanations.

State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV cache. Selective SSM with hardware-aware design. Mamba-1 (d_state=16) and Mamba-2 (d_state=128, multi-head). Models 130M-2.8B on HuggingFace.

State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV cache. Selective SSM with hardware-aware design. Mamba-1 (d_state=16) and Mamba-2 (d_state=128, multi-head). Models 130M-2.8B on HuggingFace.

Search X/Twitter profiles, tweets, trends, lists, communities, and Spaces through the AISA relay, then publish approved posts with OAuth. Use when: the user asks for Twitter/X research, monitoring, or posting without sharing passwords. Supports read APIs, authorization links, and media-aware posting.

Drafts commercial office lease agreements with rent schedules, expense allocations, maintenance splits, and protective provisions for landlord and tenant. Covers NNN, gross, and modified gross structures. Use when drafting office leases, commercial lease agreements, landlord-tenant contracts, or space rental agreements for commercial office premises.

Use when the request is broad and spans idea evaluation, MVP planning, UX, frontend, backend, database, auth, integrations, and launch strategy. Trigger on asks like "build this app", "turn this idea into a product", "design the whole system", or "plan everything end to end".

Captures a spatial hierarchy of rendered DOM elements from any webpage via Playwright CLI, returning an LLM-friendly text tree, structured JSON tree, and a node map with CSS selectors and overlay metadata.

Artistic, research-grounded UI enhancement for making an existing page feel intentional, spacious, and human-crafted. Use when "make this page nicer/polished/premium/editorial", less crowded, better laid out, typography, or visual hierarchy. Flows/AI-templated UX → enhance-web-ux. Layout contract → audit-responsive.

Design privacy-aware observability for AI agents using traces, spans, structured events, metrics, cost attribution, dashboards, alerts, and investigation workflows. Use when instrumenting an agent, debugging intermittent tool or model failures, defining service-level objectives, analyzing latency or spend, auditing agent decisions, or preparing production monitoring.

Scaffolds a per-task verification skill for a repo, including a dev-local launcher, a browser driver, and a verification SOP that spawns a sub-agent to drive the app and produce proof.

Searches and reads X (Twitter): profiles, timelines, mentions, followers, tweet search, trends, lists, communities, and Spaces. Publishes posts after the user completes OAuth in the browser. Use when the user asks about Twitter/X data, social listening, or posting without sharing account passwords.

World-class frontend engineering - React philosophy, performance, accessibility, and production-grade interfacesUse when "frontend, react, vue, svelte, next.js, nuxt, component, state management, redux, zustand, client side, spa, ssr, hydration, bundle size, web vitals, accessibility, a11y, responsive, css, tailwind, frontend, react, typescript, performance, accessibility, components, state, architecture" mentioned.

Econometrics skill for Synthetic Control Method (SCM). Activates when the user asks about: "synthetic control", "SCM", "synthetic counterfactual", "donor pool", "placebo test", "in-space placebo", "in-time placebo", "MSPE ratio", "Abadie Diamond Hainmueller", "augmented synthetic control", "penalized SCM", "synthetic DID", "合成控制", "合成控制法", "捐助池", "安慰剂检验", "合成反事实", "合成DID", "政策评估"

Deploys and manages Falco YAML rules for runtime threat detection in containers and Kubernetes, monitoring syscalls for shell spawns, file tampering, network anomalies, and privilege escalation. Parses Falco alerts for incident response.

Fetch GitHub issues, spawn sub-agents to implement fixes and open PRs, then monitor and address PR review comments. Usage: /gh-issues [owner/repo] [--label bug] [--limit 5] [--milestone v1.0] [--assignee @me] [--fork user/repo] [--watch] [--interval 5] [--reviews-only] [--cron] [--dry-run] [--model glm-5] [--notify-channel -1002381931352]

Audit Claude Code hooks for fire frequency, latency, exit codes, and output value. Surfaces hooks that never fire, hooks that fire too often (spam), hooks that exit non-zero silently, and hooks that fire but produce no output Claude actually uses. Run weekly automatically. Required after wiring new hooks.

Use this skill when the user asks what emergency equipment a CMV must carry under 49 CFR 393.95 — reflectorized triangles, flares, fire extinguishers, spare fuses, exceptions for buses/passenger CMVs, where to mount/store equipment, and what roadside inspectors check. Cite 49 CFR 393.95 + 393.97.

Routes reverse engineering, exploitation, penetration testing, malware, mobile, firmware, browser automation, documentation, and security tasks to the appropriate specialist skill. Use when a task spans modules or the correct reverse-skill entrypoint is unclear.

Trigger: macOS UI, screen recording, simulator UI, desktop app click, capture screen, system dialog, macOS clipboard, menu bar, drag drop. Scope: Automate macOS GUI applications, drive input events, list windows/spaces, and capture visual state via Peekaboo CLI. Boundary: Excludes pure browser-level automation (use chrome-devtools instead) or headless server tasks.

Builds distinctive, production-grade UI by committing to one of 12 named visual directions, implementing a real typography scale, spacing system, and one signature flourish, then gating the result through a measurable anti-slop checklist and contrast/touch-target verification.

Create, inspect, and edit PCG Graph assets — add nodes, configure settings, connect pins, and wire procedural generation graphs (native PCG Python API). Use when the user asks to build a PCG graph, add PCG nodes (Surface Sampler, Static Mesh Spawner, filters), wire PCG pins, place a PCGVolume and generate, or set up procedural scattering.

Searches and reads X (Twitter): profiles, timelines, mentions, followers, tweet search, trends, lists, communities, and Spaces. Publishes posts, likes/unlikes tweets, and follows/unfollows users after the user completes OAuth in the browser. Use when the user asks for Twitter/X research, social listening, posting, or account interactions without sharing account passwords.

Configure a CDN with optimized caching, SSL/TLS, security headers, and cache invalidation — auto-detects hosting provider and app type, generates CloudFront, Cloudflare, or Vercel config with per-content-type cache rules, edge functions, and compression. Use when deploying a static site, SPA, SSR app, or adding a CDN layer to an existing API.

Internal skill. Called by setup-evaluation after a PASS. Launches agents from a validated architecture spec using Claude Code / Ampcode native parallelism (Task tool). Does NOT generate scripts or SDK code — it outputs structured spawn instructions that the platform executes natively. Never invoked directly by the user. Never launches without a setup-evaluation PASS.

Layer 4 of the keyword research pipeline. Spawns a "skeptical SEO" adversarial sub-agent to argue against every survivor of Layers 1-3. Catches mechanical-classifier blind spots — wrong SERP intent, hidden link-graph gauntlets, AIO trajectory shifts, vanity-rank metrics. Same pattern as quality-check's adversarial draft read, applied to keyword selection.

Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limited compute (5× cost reduction vs dense models), implementing sparse architectures like Mixtral 8x7B or DeepSeek-V3, or scaling model capacity without proportional compute increase. Covers MoE architectures, routing mechanisms, load balancing, expert parallelism, and inference optimization.

Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning neural networks, debugging loss spikes or OOM, choosing architectures, or optimizing GPU throughput.

Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limited compute (5× cost reduction vs dense models), implementing sparse architectures like Mixtral 8x7B or DeepSeek-V3, or scaling model capacity without proportional compute increase. Covers MoE architectures, routing mechanisms, load balancing, expert parallelism, and inference optimization.

Searches and reads X (Twitter): profiles, timelines, mentions, followers, tweet search, trends, lists, communities, and Spaces. Publishes posts, likes/unlikes tweets, and follows/unfollows users after the user completes OAuth in the browser. Use when the user asks about Twitter/X data, social listening, posting, or interacting with tweets/users without sharing account passwords.

Builds complete bilingual email drafts from intent and audience. Use when the user asks to write, rewrite, localize, polish, or generate an email, campaign email, outreach message, customer update, follow-up, announcement, invite, sales email, internal note, or lifecycle email and wants clear subject lines, preheaders, body copy, CTAs, variants, or English and Spanish versions.