Total 52,658 skills, AI & Machine Learning has 8822 skills
Showing 12 of 8822 skills
Analyze medication adherence and management platforms including dose tracking accuracy (MPR, PDC metrics), drug-drug and drug-food interaction checking completeness, refill prediction algorithms, dosage schedule optimization with conflict detection, caregiver notification escalation workflows, pharmacy system integration (NCPDP, HL7 FHIR), adverse event signal detection, smart dispenser integration, and alert fatigue mitigation for patient safety systems.
Use when building durable AI agents or agentic workflows with Inngest and AgentKit, including model calls, tool calls, multi-agent networks, human approval, realtime progress, provider rate limits, crash-safe execution, and Agent Evals handoff. Covers AgentKit, `step.ai`, `step.run`, `step.waitForEvent`, native realtime, and when to use lower-level Inngest primitives instead of an in-memory agent loop. Use `inngest-agent-evals` with this skill when the user wants scoring, sessions, experiments, deferred scorers, or outcome-based evaluation for the agent.
Build and flash the XIAO ESP32S3 Sense camera web apps in STA (router) mode: a Teachable-Machine-style dataset collector page and a live inference viewer page, reachable at http://<name>.local while every device KEEPS its internet connection. Use this skill whenever the user wants the camera web app on their normal WiFi network — home/office development, "인터넷 안 끊기게", "공유기로", "mDNS", or says "STA 모드". For router-less classroom hotspot deployments use the xiao-webcam-ap skill instead.
Use when building, migrating, or debugging Agent Evals on Inngest: scoring AI agent or workflow outcomes, deferred scorers, sessions, traces, step experiments, experiment variant attribution, Insights queries, or production eval loops for prompts, models, tools, providers, and agent behavior. Covers TypeScript SDK v4 scoring beta APIs, `scoreMiddleware`, `step.score`, `inngest.score`, `createScorer`, `defer`, `group.experiment`, `experimentRef`, `meta.sessions`, and when to use durable workflow primitives for outcome-based evaluation.
Build and flash the XIAO ESP32S3 Sense camera web apps in STANDALONE AP (hotspot) mode: a Teachable-Machine-style dataset collector page and a live inference viewer page served by the board itself at http://192.168.4.1. Use this skill whenever the user wants the camera web app WITHOUT a router — classroom/education deployments, demos with no WiFi, per-student boards, or says "AP 모드", "핫스팟", "공유기 없이". For router (STA) mode use the xiao-webcam-sta skill instead.
Train and deploy a TinyML model for the XIAO ESP32S3 (Sense) using the Edge Impulse REST API only — no edge-impulse-cli needed (its serialport dep fails to build on modern Node/Windows). Covers: dataset upload, impulse creation (audio MFCC / vision transfer-learning), training jobs, downloading the Arduino library, and the on-device fixes required to actually run it on the ESP32-S3. Use this skill whenever the user wants to train/retrain a model ("재훈련", "edge impulse", "TinyML 훈련", "모델 배포"), upload a dataset to Edge Impulse, or gets EI Arduino-library build/runtime errors (mel filterbank, objs.a, tensor arena, EI_MAX_OVERFLOW_BUFFER_COUNT).
Turn a vague feature or product idea into an agreed, persisted specification through relentless structured questioning. Never assumes — every gap, ambiguity, or "probably" becomes a question to the developer, and the spec cannot be approved while open questions remain. Produces docs/specs/<NNN>-<slug>.md with acceptance criteria that /plan, /scaffold, and /tdd consume. Use when: "spec", "write a spec", "spec this out", "requirements", "PRD", "acceptance criteria", "define the feature", "user stories", "what should we build", or before planning any feature too big to describe in one sentence.
Owns the session handoff lifecycle: the end-of-session ritual that captures completed work, pending tasks, and learnings into .claude/handoff.md, and the session-start protocol that loads it back. Triggers on: /wrap-up, "wrap up", "done for today", "that's all", "end session", "signing off", "handoff" — and at session start: "start session", "session start", "load handoff", "pick up where we left off", "what were we working on".
Grounds a coding agent's architecture decisions in real arXiv prior art before it builds something new. Reads arXiv category-wise via real HTTP fetch, spawns parallel isolated reads across the papers found, scores/clusters them, then converges on ONE recommended path with citations, a first step, and known prior-art pitfalls to avoid. Use on /neuroarxiv, before designing non-trivial architecture, algorithms, ML/systems techniques, or protocols, or when the user asks "has anyone solved this", "what's the state of the art", or "am I about to rebuild something that already exists". Skip for trivial CRUD, glue code, or closed phrasing ("just", "quick", "standard"). Full pre-flight gate is in the skill body.
任意の対象リポジトリに Claude Code の .claude/ 体系（CLAUDE.md・Agents・Rules・Skills・hooks）を 初期セットアップする。「claude セットアップして」「.claude 作って」「CLAUDE.md 初期化」「Agent 整備して」 「claude-code セットアップ」などで使用。既存 .claude/ の差分充実は update-claude を使用。 implement-issue-tree が動く前提（gh auth / sub_issues / workflow js）の整備まで含む。
Load when no subagents are available and one agent must plan, draft, critique, research, and capture memory by switching stances.
Use this skill when the user wants to turn a recurring browser workflow into a reusable, parameterized agent skill, especially when the task has concrete inputs and a clear output such as scheduled scrapes, form submissions, data extraction, monitoring flows, price probes, or login-gated reports. Do not use for one-off web tasks; use steel-browser.