Deploy and operate the RTVI-CV-3D microservice as MV3DT (`MODE=mv3dt`): per-camera DeepStream perception plus BEV Fusion over calibrated cameras. Supports the bundled sample dataset, custom video files, and RTSP streams, and chains to `vss-generate-video-calibration` when calibration is missing. Use `vss-deploy-profile` for the full warehouse blueprint and `vss-deploy-detection-tracking-2d` for single-camera 2D detection.

深度调研 - 当需要检索真实世界细节、推演复杂逻辑、获取风格参考时触发。关键词：调研、检索、搜索、参考、素材。

Configure `MaterialApp.router` using a package like `go_router` for advanced URL-based navigation. Use when developing web applications or mobile apps that require specific deep linking and browser history support.

Zigbee, Thread, and Matter mesh-protocol attack methodology — IEEE 802.15.4 sniffing with TI CC2531 / CC2540 / Sonoff Zigbee Dongle E, KillerBee toolkit, Touchlink commissioning abuse with the well-known transport key, replay/injection attacks, Zigbee Cluster Library command abuse for door locks and bulbs, Thread network credential theft, Matter commissioning chain analysis, and 6LoWPAN/IPv6 routing exploitation. Use when targeting smart-home or commercial mesh deployments, Zigbee-based door locks, lighting, or sensor networks.

Comprehensive OSINT methodology skill for offensive security, red team intelligence gathering, and bug bounty reconnaissance. Covers domain recon, email harvesting, social media profiling, GitHub/code leaks, Shodan/Censys enumeration, breach data lookup, employee profiling, infrastructure mapping, cryptocurrency tracing, geospatial intelligence, and AI-assisted analysis workflows. Use when performing reconnaissance against a target domain or organization, investigating a person or entity, tracing cryptocurrency flows, geolocating images or events, or building an attack-surface map.

Set up recommended workspace preferences for new team members. Use when you are starting in a new workspace and want guided defaults applied.

Use twitter-cli for ALL Twitter/X operations — reading tweets, posting, replying, quoting, liking, retweeting, following, searching, user lookups. Invoke whenever user requests any Twitter interaction.

Integrate markstream-vue2 into a Vue 2.6 or 2.7 app. Use when Codex needs Vue 2-compatible setup, `@vue/composition-api` decisions, CSS wiring, optional peer setup, or scoped Markstream overrides in a Vue 2 repository that is not specifically Vue CLI / Webpack 4 constrained.

用于 Python 后端面试出题；覆盖 Python 基础、数据库、Django/Flask、缓存与部署，强调工程落地。

用于阿里后端面试出题；强调业务深度、项目落地和中间件选型能力，关注复杂场景下的取舍与闭环。

Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

File upload endpoints in ASP.NET minimal APIs (.NET 8+)

Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers.

Protein structure prediction from sequence. ESMFold-based, single GPU, no MSA needed. Predicts 3D structures with pLDDT confidence scores for drug discovery targets.

Strict SMILES validation, structural comparison, and modification verification. Catches invalid LLM-generated molecules.

Retrieve structurally similar compounds with known properties from ChEMBL/ZINC to ground predictions and inform optimization. Based on MolRAG (Xian 2025, ACL).

End-to-end molecular docking pipeline. Target preparation, pocket detection, protein-ligand docking (DiffDock/Vina), scoring, interaction analysis, and pose ranking.

De novo molecule generation for drug discovery. Scaffold-based analog enumeration, fragment growing/linking, structure-based design, multi-objective optimization, and drug-likeness filtering.

Interpretable ADMET analysis with mechanistic reasoning. Maps liabilities to structural causes and biological pathways. Based on CoTox (Park 2025) and DrugR (Liu 2026).

Semantic search over ingested documents using RAG (LlamaIndex/ChromaDB or Foundational RAG)