household-review-pack

Finds the clients an advisor is quietly losing from their own book, ranking the households that have gone silent past their usual rhythm by how they felt last time, whether their sentiment was already falling, whether they stopped driving the conversation, and whether they have named another advisor. Says who to call first and how to open. Use whenever the user asks who is at risk, who has gone quiet, who they haven't spoken to in a while, who to call this week, or whether they are losing anyone — and for a re-engagement list, a brief on one household, or a draft re-engagement email. Requires Zocks MCP — will not run from transcripts, uploads, or any other source.

Reads one advisory client or household two ways — a fast temperature check on whether they are cooling, steady or warming, and a full behavioral profile covering what motivates them, what they are afraid of, how they decide, whether their stated risk tolerance matches their behaviour, and how engaged they have been — then a calibrated stance for the next conversation. Use whenever the user asks what a client is like, how to approach them, what motivates them, how a meeting went, what the mood is, why they won't commit, or whether a relationship is cooling or warming. Requires Zocks MCP — will not run from transcripts, uploads, or any other source.

Builds a client-ready Client Journey for one household — a five-stage lifecycle timeline marking where the family sits today, a relationship tree including beneficiaries never met, their outside professional team, a dated timeline of life events and turning points, and every motivator, worry and consequential fact on record with its corroboration and age — plus what they have never been asked that the rest of the book covers. Use whenever the user wants everything known about a household, a family tree or lifecycle report, who is who in a family, what is new on a client, or a comprehensive picture polished enough to walk a client through. Requires Zocks MCP — will not run from transcripts, uploads, or any other source.

Ranked pipeline of held-away asset opportunities from the advisor's own recent meetings — old 401(k)s, outside brokerage accounts, inheritances, business sale proceeds, real estate, spouse assets — grouped by household and marked new, raised again, nothing new since, task created, or client agreed. Use whenever the user asks what held-away or outside money came up recently, wants their consolidation pipeline or money in motion, mentions old 401(k)s or outside accounts, or says things like "run the radar" or "held-away opportunities". Requires Zocks MCP — will not run from transcripts, uploads, or any other source.

Deep single-household analysis naming the one highest-value next action and proving why — unused tax space, a closing Roth conversion window, RMD and IRMAA sequencing, Social Security and survivor math, irreversible pension elections, tax drag, protection and estate gaps. Never the advisor's own open tasks read back. Shows the runner-ups it rejected and names who outside the firm has to act. Use whenever the user asks what to do for a client, the best move, the biggest opportunity, what they are missing with someone, or wants a household's situation analysed. For a household's full tax picture, tax-opportunity-scan is the skill. Requires Zocks MCP — will not run from transcripts, uploads, or any other source.

Builds a household's full tax opportunity picture from every analyzed meeting on record, across two horizons — current-year items with deadlines (deferral capacity, a second plan nobody checked, taxable-account drag, giving structure, equity dates, education vehicles) and multi-year strategies 3-5 years out (conversion windows, drawdown sequencing, the survivor's position, business structure, property sales). Ranks by impact, cites the meeting each item came from, names the tax questions nobody has asked, and flags what the household raised that went nowhere. Use whenever the user asks about tax planning, tax opportunities or tax exposure for a client, wants a tax review or CPA handoff, asks what they are missing on someone's taxes, mentions Roth conversions, deferral capacity, tax drag, harvesting, charitable giving or equity comp timing, or asks what to cover before year end or tax season. Advisor intelligence, not advice. Requires Zocks MCP — will not run from transcripts, uploads, or any other source.

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".

Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.

This skill should be used when the user wants to "write agent code", "build an agent with ADK", "add a tool", "create a callback", "define an agent", "use state management" — in a project that needs ADK (Agent Development Kit) API patterns and code examples. It provides a quick reference for agent types, tool definitions, orchestration patterns, callbacks, state management, the graph Workflow API, and reference recipes to study. Do NOT use for scaffolding (use google-agents-cli-scaffold) or deployment (use google-agents-cli-deploy).

Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.

Generate AI music on RunComfy via the `runcomfy` CLI — a smart router across the music-model catalog. Routes to ElevenLabs AI Music Generation (premium 44.1 kHz stereo vocal tracks, 5 s–5 min, $0.0083/s) and ACE Step / ACE Step 1.5 (StepFun-AI open-weights, tag-driven composition, multilingual lyrics, $0.0002–0.0003/s, ~27× cheaper), plus ACE Step audio-inpaint (regenerate a time range inside an existing track) and ACE Step audio-outpaint (extend a track before or after). Picks the right model for the user's actual intent — premium vocal hook, cheap background music library, multilingual pop song, repair a bad chorus, lengthen a 30 s draft into a 2 min cut — and ships each model's documented prompting patterns plus the minimal `runcomfy run` invoke. Triggers on "generate music", "make a song", "AI music", "background music", "instrumental track", "soundtrack", "jingle", "theme music", "royalty-free music", "compose", "music with lyrics", "extend music", "fix this song", "inpaint music", or any explicit ask to generate or edit music.

创建 lark-cli 的自定义 Skill。当用户需要把飞书 API 操作封装成可复用的 Skill（包装原子 API 或编排多步流程）时使用。