k1064190

Morning market briefing with stock picks for US and Korean markets. Generates a daily report covering macro environment, sector rotation, 10-12 actionable predictions (5-6 per market) with BUY/WATCH/HOLD/AVOID/SELL labels and Korean reasoning, plus an auto-Toss-synced portfolio review section recommending hold/add/trim/exit per current position. Each pick is logged as a formal prediction for track record tracking. Triggers on keywords like daily briefing, morning report, market overview, today's picks, what should I trade, 오늘 시장, 일일 브리핑.

Stock expectation analysis. Combines algorithmic technical scoring + structured news/sentiment + LLM macro/narrative context + multi-horizon directional predictions to emit a deterministic BUY / WATCH / HOLD / AVOID / SELL recommendation per stock. Modes: (1) /expect KR or /expect US auto-discover 5 trending stocks, (2) /expect NVDA or /expect 삼성전자 single-ticker deep dive, (3) /expect NVDA,AMD,AVGO multi-ticker batch, (4) /expect ALL or /expect runs both markets. Triggers: expect, expectation, stock picks, what should I buy, hot stocks, trending stocks, buy or sell, 기대값, 종목 추천, 뭐 사야돼, 매수, 매도, 핫한 종목

Monthly ISA contribution briefing for the long-term Korean ETF book. Values the ISA portfolio, checks drift vs the approved target allocation, optionally proposes a tilt (code-clamped to ±10%p), runs the sell-free contribution allocator, and composes a Korean briefing with per-ETF buy amounts, track record vs benchmarks, and the decision-log id. Triggers on keywords like isa, ISA 브리핑, 적립, 월 적립, 리밸런싱, ETF 적립, isa briefing, monthly contribution.

Unified forward catalyst timeline + deterministic R3 event-risk gate. Merges the FMP earnings calendar (US-listed) and economic calendar (FOMC/CPI/NFP) into one timeline, then turns imminent binary events into label caps and confidence trims so /expect and daily-briefing don't issue fresh BUY calls right into an earnings print or a Fed decision. Triggers: catalyst, event risk, earnings risk, event gate, upcoming earnings, FOMC, CPI, NFP, macro event, 이벤트 리스크, 실적 발표 임박, 캐털리스트.

Set and monitor price-level alerts for a saved watchlist plus open predictions and portfolio positions. Fires Korean Telegram alerts when the latest close touches an entry zone, stop, target, or re-entry level. DELAYED / EOD-ish, not real-time. Use when the user wants to track entry/exit levels, set an alert, or watch tickers. Triggers on keywords like watchlist, alerts, set alert, entry zone, 워치리스트, 관심종목, 알림.

Decide when to exit, trim, add to, or hold each current holding. Layers an ATR chandelier trailing stop, R:R take-profit, and linked-prediction thesis invalidation on top of the portfolio. Use when the user asks to check positions, where to set stops, when to take profit, or whether to sell. Triggers on "포지션 점검", "손절", "익절", "트레일링 스탑", "언제 팔아", "exit check", "trim", "take profit", "when to sell", "stop loss", "trailing stop".

This skill should be used when analyzing sector rotation patterns and market cycle positioning. It fetches sector uptrend data from CSV (no API key required) and optionally accepts chart images for supplementary analysis. Use this skill when the user requests sector rotation analysis, cyclical vs defensive assessment, overbought/oversold identification, or market cycle phase estimation. All analysis and output are conducted in English.

Prescriptive sector-rotation screener that ranks sectors by relative strength, constituent breadth, and lifecycle stage, emitting a machine-readable FAVOR / ROTATING_IN / ROTATING_OUT / AVOID / NEUTRAL verdict per sector for both US and KR markets. Unlike the narrative sector-analyst skill, this produces a deterministic JSON snapshot (data/sector_rs_{market}.json) that the daily-briefing candidate discovery consumes to bias stock ranking toward sectors rotating in and away from sectors rotating out. Use when the user wants an actionable "which sectors to favor/avoid right now" answer, a sector leaderboard, sector relative-strength ranking, 섹터 로테이션, 섹터 상대강도, or which sectors are leading/lagging.

Delegate tasks to Google Antigravity CLI (`agy`) running as a subagent. Use this skill whenever the user says "use antigravity", "ask agy", "run this with antigravity", "delegate to agy", wants a second opinion from a different AI model, needs parallel AI execution, or wants to offload an agentic task to a separate process. Antigravity CLI is Google's successor to the Gemini CLI (Gemini CLI's free tier was retired 2026-06-18); it exposes Gemini 3.x, Claude Sonnet/Opus 4.6, and GPT-OSS models behind the single `agy` command. Also trigger when the codebase or files are too large for Claude's context window and Antigravity's large context is needed for full codebase analysis.

Use when the user asks to run Claude Code CLI non-interactively (claude -p, "spawn a claude instance", "delegate to claude", "use a fresh claude context") or when an isolated Claude Code run is needed for parallel work, scoped tool permissions, or a clean-context second pass. Also use when a Codex or Gemini/Antigravity orchestrator needs Claude's capabilities for a subtask — this skill has the exact command patterns.

Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex for code analysis, refactoring, or automated editing

Delegate tasks to Gemini CLI running as a subagent. Use this skill whenever the user says "use gemini", "ask gemini", "run this with gemini", "delegate to gemini", wants a second opinion from a different AI model, needs parallel AI execution, or wants to offload a long-running agentic task to run in the background while Claude continues working. Also trigger when the user wants to leverage Gemini's built-in Google Search, wants to compare results from two different AI models, or when the codebase or files are too large for Claude's context window and Gemini's massive context is needed for full codebase analysis.

Use this skill whenever the user mentions Google's Gemini in conjunction with unofficial, reverse-engineered, web-based, cookie-based, or free access methods. Also triggers for any implementation of Gemini using browser cookies, __Secure-1PSID authentication, or when the user wants to access Gemini's web API programmatically without using the official Google AI SDK. This includes requests for "free Gemini", "unofficial Gemini API", "Gemini without API key", "access Gemini via web", or similar patterns.

Generates a structured retrospective document from the current conversation context and user's free-form reflections. Use this skill whenever the user wants to reflect on their work session, write a retrospective, capture lessons learned, or review what happened during a coding session. Trigger on any mention of retrospective, reflection, review of work done, postmortem, or session wrap-up — even if the user just says "let's reflect" or "what did we learn today."

Cut a new gemini-voyager release — open-issue triage, preflight checks, version bump, 10-locale changelog, commit, tag, push, curated GitHub release body, Edge store zip, and Safari DMG. Use whenever the user says "发版", "release", "bump", "cut a release", "ship vX.Y.Z", or otherwise signals shipping a new version. Also use when the user wants just an Edge zip or Safari DMG for an existing release.