Total 56,386 skills, AI & Machine Learning has 9387 skills
Showing 12 of 9387 skills
Writes, rewrites, diagnoses, and improves any LLM prompt with minimal, high-signal edits. Use when the user wants to create a new prompt from scratch, review or fix a prompt that produces poor output, simplify or tighten instructions, restructure a long prompt, port a prompt between models, or expand an existing prompt. Covers system prompts, agent instructions, CLAUDE.md rules, SKILL.md prompt bodies, chat templates, structured-output prompts, RAG context templates, and prompt strings embedded in code. Also use when editing any file whose primary content is LLM instructions.
Optional AI SDLC architecture workflow. Use when an AI assistant needs to define system boundaries, components, interfaces, architectural constraints, alternatives, decisions, tradeoffs, risks, or validation for a feature and produce routed human and machine artifacts linked to requirements and durable decisions. Supports `--quick-flow` for focused design and `--full-flow` for strict decision, risk, and validation coverage.
AI SDLC controlled change-workspace and specification-delta workflow. Use when an AI assistant needs to create or validate an isolated proposal workspace, author and validate requirement deltas, preview canonical changes, or apply and archive an explicitly approved change with rollback evidence. Supports `--quick-flow` for assumption-driven drafts and `--full-flow` for strict owner, target, evidence, and authority checks.
AI SDLC package trust and privacy-preserving local metrics workflow. Use when an AI assistant needs to verify package origin, file integrity, harness compatibility, declared capabilities, provenance evidence, or generate reproducible aggregate run, retry, budget, coverage, and freshness metrics without collecting source, prompts, commands, or diffs. Supports `--quick-flow` and `--full-flow`.
AI SDLC reusable quality-lens workflow. Use when an AI assistant needs to challenge a requirement, design, plan, test strategy, change, or delivery artifact through pre-mortem, adversarial, edge-case, stakeholder-conflict, reversibility, abuse-case, operational-failure, or assumption lenses and finalize evidence-backed findings with ownership and traceability. Supports `--quick-flow` for selected high-value lenses and `--full-flow` for the complete applicable registry.
Designs, builds, and deploys AI agents or multi-agent systems on Google Cloud. Provides an interactive workflow to gather requirements, recommend a tailored architecture, and generate deployment instructions. Use when designing or implementing agentic systems on Google Cloud. Don't use for general Google Cloud solution architecture (use google-cloud-solution-architecture instead) or for narrow tasks targeting a single product without agent context.
Discovers requirements, and generates architectural, design, and deployment guidance for a retrieval-augmented generation (RAG)-capable enterprise search system in Google Cloud. Use when users need a vector-enabled SQL database as the store and index for the embedding vectors, an open model and open-source inferencing framework, and Kubernetes containers to host all the application components. DON'T use this skill for fully-managed RAG, or SaaS search services, or when a non-SQL vector database is required.
Tool-neutral CLI agent rules for TI MSPM0 development with Code Composer Studio, Keil/uVision, CMake/GCC/OpenOCD, SysConfig, and DriverLib. Use when an agent needs to inspect or modify MSPM0 projects, edit .syscfg configuration, avoid generated SysConfig/build files, use DriverLib APIs, validate SysConfig output, package reusable MSPM0 examples, or work on NUEDC-style MSPM0 embedded firmware.
Parallel DAG-plan implementation skill. Reads a v-planning plan directory (root.md + step-<n>.md files), topologically schedules ready steps, and fans them out as parallel sub-agents. Use whenever the user invokes /v-implement, points at a plan directory produced by /v-plan, or asks to "run the parallel plan", "implement the DAG", or "fan out the steps" — even without those exact words. For linear plans (single .md file), use `implementing` instead.
A creative-direction (taste) layer for music videos and short-form edits in the angelcore / cloud-trance / hyperpop visual family. Distills a named-genre aesthetic vocabulary, a mood + color + light system, and a beat-synced editing grammar, then chains ECC's video skills (video-editing, fal-ai-media, remotion-video-creation, motion-*, content-engine) into one production pipeline. Use when the work is not just making a video function but making it feel intentional, when building a music video, a fancam/edit, a moodboard-driven reel, or when choosing a coherent visual direction for AI-generated b-roll.
Run an ML training job on a completed Itô compute booking through the canonical Itô backend. Use after ito-compute has booked GPU nodes and the user wants pre-training, fine-tuning, or RL on that metal. Chains off a booking record; ECC implements no training stack of its own.
Use when writing DALI data loading or preprocessing code with `nvidia.dali.experimental.dynamic` (ndd), or when converting DALI pipeline-mode code to dynamic mode, or when the user asks about DALI dynamic mode, imperative DALI, or ndd. Use this skill any time someone mentions 'ndd', 'dynamic mode', or wants to load/augment data with DALI outside of a pipeline definition.