菜单
这个仓库中的 skills
SkillsMP 已收集 DreamLab-AI/agentbox 中的 161 个 Skill。打开任一 Skill 可查看来源和详情。
DreamLab-AI/agentbox已展示 40 / 161 个已收集 Skill。
Browser automation, web scraping, visual testing, and WebGPU validation via the browsercontainer sidecar (chrome-devtools-mcp, 40+ tools). Use for navigating pages, clicking elements, filling forms, taking screenshots, executing JavaScript, accessibility…
原文语言：英语
Train and fine-tune deep learning models in PyTorch with CUDA GPU acceleration — nn.Module definition, training loops, DataLoaders, checkpointing, mixed-precision (AMP), and transfer learning with torchvision/torchaudio pretrained models. Use when writing…
原文语言：英语
Design AI-first interfaces around ongoing user relationships — persistent memory, graduated trust, and human+AI collaborative planning — instead of isolated screens. Use for agentic/relationship-centric UX work: systems that learn and remember across…
原文语言：英语
Source-grounded answers about the RuvNet ecosystem — ruflo, ruvector, safla, agentdb, agentic-flow, sparc and ~21 sibling repos. Use whenever a task asks how a RuvNet tool works, or before you reach for a generic alternative (Pinecone, LangChain, ChromaDB,…
原文语言：英语
Generate Minecraft Java Edition worlds from real-world geographic data. Converts OpenStreetMap buildings, roads, water, and terrain into playable Minecraft worlds using the arnis Rust engine. Integrates with QGIS for geospatial analysis, Blender for 3D…
原文语言：英语
Full SEO audit: Google Search Console data + technical crawl + search intent analysis. Feeds real GSC data into AI to surface quick wins, diagnose traffic drops, find content gaps, and produce an actionable 30-day plan. Use this skill whenever the user asks…
原文语言：英语
AI-powered SEO/SEM automation suite with 6 specialised skills. Google Search Console integration, content writing (E-E-A-T), keyword research, meta tag optimisation, JSON-LD schema markup, and GEO (Generative Engine Optimisation) for AI citations. Use when…
原文语言：英语
Upgrade toprank skills to the latest version. Detects global vs vendored install, runs the upgrade, and shows what's new. Use when asked to "upgrade toprank", "update toprank skills", or "get latest version". Also handles inline upgrade prompts when a skill…
原文语言：英语
UI/UX design intelligence for web and mobile: recommends styles, colour palettes, font pairings, layouts, and chart types, plus stack-specific implementation guidance (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui). Use when…
原文语言：英语
Comprehensive Wardley mapping toolkit that transforms any input (structured data, unstructured text, business descriptions, technical architectures, competitive landscapes, or abstract concepts) into strategic Wardley maps. Default render path is **Mermaid…
原文语言：英语
High-performance WebAssembly graphics with JavaScript interoperability. Use when building performance-critical web graphics, real-time animations, computational geometry, or hybrid JS/WASM architectures.
原文语言：英语
Multi-source web research via the web-researcher-mcp Go server (deployed binary v1.43.0) — you pick the search ENGINE and the trusted SOURCES, and every citation is a real, checkable link. 21 live MCP tools (verified against tools/list):…
原文语言：英语
Summarise a single web page or YouTube video into short/medium/long notes and extract semantic topic links for Logseq or Obsidian. Trigger when the user says "summarise this URL/article/video", "get the YouTube transcript", "pull the key points from this…
原文语言：英语
Bulk backfill markdown transcript files for a YouTube podcast series. Downloads transcripts, show notes, and links for all episodes in a date range, then runs source extraction and applies enrichment tables. Designed for one-off historical backfill — for…
原文语言：英语
Batch-download YouTube channel transcripts, show notes, and links as markdown files. Enriches each file with source extraction from transcripts, cross-checked via Haiku, and optionally downloads referenced reports/papers to a local assets directory. Designed…
原文语言：英语
Raw Chrome DevTools Protocol scripting against a live Chrome session — direct WebSocket, no Puppeteer. Use when you need low-level CDP access: WebGPU/WebGL GPU testing, console/JS evaluation, screenshots, or performance traces on the browsercontainer GPU…
原文语言：英语
Ground agent reasoning in DreamLab's formal knowledge graph (5,975 OWL classes, Oxigraph/Whelk) via the pervasive ontology binding (PRD-020/ADR-112). Use when you want to ground or augment thinking in the ontology, check what the knowledge graph says about a…
原文语言：英语
Calibrate response style to the user's communication register — transactional/direct vs relational/high-context — and decide whether to answer, explore, or ask a clarifying question first. Use when a request is ambiguously framed: hedging ("maybe", "wondering…
原文语言：英语
Lock-free version control for multiple AI agents committing concurrently to the same repo, built on Jujutsu (jj). Use when several agents need to commit/branch/rebase simultaneously without lock contention, for conflict-free concurrent worktree isolation, or…
原文语言：英语
AEC (Architecture, Engineering, Construction) studio for building and site work. Use when designing a building, planning or assessing a site, running NYC zoning or property due diligence, programming workplace space from headcount, comparing material…
原文语言：英语
Collaborative UI/UX design guidance for building unique, accessible, non-generic web interfaces. Use when building or styling web components, pages, or apps and making visual decisions about colour, typography, layout, motion, or accessibility — especially…
原文语言：英语
Router for browser-automation tasks — picks the right tool and points to the canonical sidecar setup. Use when driving a real browser: navigating pages, clicking, filling forms, taking screenshots, reading console or network traffic, debugging a web UI, or…
原文语言：英语
Drive a real Chrome via the browsercontainer sidecar over MCP SSE. Use when a task needs a live browser — navigate, click, fill forms, screenshot, read the accessibility tree, run JavaScript in-page, or validate WebGPU/WebGL rendering. Canonical owner of the…
原文语言：英语
Local-GPU video understanding and editing via MCP — analyse footage, find the best moments, cut highlight reels, add captions, render platform-ready clips (TikTok/Reels/ Shorts/YouTube), clone voices, lip-sync avatars, generate music, and run a real-time…
原文语言：英语
Structural code-intelligence MCP for large codebases — trace call graphs, get architecture overviews, score git-diff impact, search symbols, and manage ADRs from a persistent tree-sitter knowledge graph. Use on large or unfamiliar repos (500+ files) for "what…
原文语言：英语
Systematic visual UI/UX audit that produces phased, implementation-ready design plans, plus focused single-objective refinement passes ("lenses"). Use when the user asks to audit a UI, improve or polish an app's visual design, review design consistency, fix…
原文语言：英语
Validate and align a whole documentation corpus against the codebase — broken-link/orphan detection, Diataxis structure, front-matter metadata, Mermaid diagram compliance (ASCII→Mermaid), UK-English spelling, and navigation/index generation — optionally via…
原文语言：英语
Train distributed neural networks (feedforward, LSTM, GAN, transformer) in E2B cloud sandboxes via Flow Nexus, no local GPU. Use when a task needs cloud-fleet distributed NN training via Flow Nexus. Not installed here (needs a Flow Nexus account); for local…
原文语言：英语
Manage Flow Nexus cloud platform resources — authentication, E2B sandboxes, app/marketplace deployment, credits & billing, coding challenges. Use when a task targets a Flow Nexus account. Requires credentials; not installed here by default.
原文语言：英语
Deploy and coordinate cloud AI swarms on Flow Nexus infrastructure (event-driven workflows, message-queue processing, scale-out agent coordination). Use when a task calls for cloud-hosted swarm state beyond local resources. Gated: needs a Flow Nexus account…
原文语言：英语
Game development studio for Godot (native), Unity, and Unreal projects — design, programming, art, audio, QA, production, and multi-agent team orchestration. Use when starting or building a game, adding a gameplay feature spanning multiple systems, running…
原文语言：英语
Google Cloud CLI (gcloud/gsutil/bq) for GCP operations from agentbox: Compute Engine VMs, Identity-Aware Proxy (IAP), Secret Manager, Cloud Run, Artifact Registry, Cloud NAT/firewall. Use when a task involves deploying to or operating Google Cloud — standing…
原文语言：英语
Multi-agent code review for GitHub PRs. Use when reviewing a pull request, running security/performance/architecture/style checks on a diff, or coordinating specialized review agents and posting findings back to a PR.
原文语言：英语
Coordinate work across many GitHub repositories at once — org-wide dependency/security updates, package and doc version alignment, cross-repo refactors, and shared architecture/templates. Use when a change must land in several repos together, not for…
原文语言：英语
Orchestrate GitHub releases end-to-end: version bumps, changelog/release-note generation, multi-platform builds, staged deployment, and rollback. Use when cutting a release, generating a changelog, coordinating a multi-package or multi-repo release, or…
原文语言：英语
Automate GitHub Actions workflows and CI/CD pipelines with claude-flow swarm coordination. Use when creating, optimizing, or debugging GitHub Actions workflow YAML, CI/CD pipelines, or GitHub repository automation — not for PR code review, releases, issues,…
原文语言：英语
Automate coordination, formatting, and learning around Claude Code operations with hooks. Use when setting up pre/post task hooks, session handoffs between agents, Git commit gating, memory coordination, or neural pattern training.
原文语言：英语
Systematic architectural thinking for the decisions AI can't own — domain modeling, systems thinking, constraint navigation, AI-aware decomposition. Use when facing system design, technology choices, integration or breaking-change planning, or any decision…
原文语言：英语
Programmatically build and run Jupyter .ipynb notebooks end-to-end via the jupyter-notebooks MCP server — add/delete/move/update cells, execute cells or whole notebooks, manage kernel state, capture outputs, and export to HTML/PDF/Python. Use when authoring…
原文语言：英语
Compile LaTeX documents to PDF - academic papers, theses, Beamer presentations, technical docs with math, bibliographies, and multi-file projects. Use when writing or compiling .tex/.bib source, typesetting equations, or building Beamer slides.
原文语言：英语