Once you have connected Claude to the Zocks MCP Server, Claude can read your analyzed meetings and answer from them — everything discussed with a household on a topic over any period, a client's motivators and concerns, sentiment and engagement trends, planning gaps, meeting scores, referral opportunities, compliance and suitability flags — with every answer sourced back to the meeting it came from.

The Zocks Advisor Intelligence plugin goes one step further: it adds seven named jobs to Claude, each one a complete piece of analysis that runs the same way every time. You pick the job. Claude does the rest.

This article covers what plugins and skills are, what is in the plugin, how to add it, how to run a skill, and what each of the seven skills does.

Three words worth separating:

The practical difference: a prompt gives you whatever you thought to ask for that day. A skill gives you the same thorough analysis every time, from anyone on your team, whether or not they are good at writing prompts.

Seven skills, all built on your own Zocks meeting data:

This is a one-time step. The seven skills are then available in chat and in Cowork.

Two ways, and they do the same thing:

At the start of a run you choose the shape of the output: an interactive report you can filter, sort and expand, or a document. Both carry exactly the same content, and every fact in either links back to the meeting it came from in Zocks.

If Claude asks you a question before it starts — which household you meant, how far back to look — answering takes a second and makes the result better. Where several contacts match a name, it asks rather than guessing.

Some skills accept extra commands while they are running — those are listed as follow-ups below. You type them in the same chat once the report is on screen.

Start it: /held-away-radar — or "what outside money came up recently?"

What it does: scans your recent meetings for outside assets — old 401(k)s, brokerage accounts elsewhere, an inheritance, business sale proceeds, property, a spouse's accounts — and groups them by household.

What you get: a ranked pipeline, each item marked new, raised again, nothing new since, task created, or client agreed, with the meeting behind it — plus how often an outside asset came up, how often you suggested a move, and how often the client agreed.

Use it: Friday afternoon, so next week's consolidation conversations are picked before Monday.

Start it: /tax-opportunity-scan — or "what are we missing on the Morgans' taxes?"

What it does: builds one household's tax picture from every analyzed meeting on record, across two horizons — current-year items with deadlines, and multi-year strategies three to five years out.

What you get: the opportunities ranked and cited to their meeting, what this household raised themselves that went nowhere, how thin or old the record is across the nine planning domains, and the tax questions nobody has asked them yet.

Follow-ups: gaps for the open questions on their own, detail [item] to expand one opportunity, handoff for a note to their CPA.

Use it: before a year-end review, a plan meeting, or a CPA handoff. Bring their tax return if you have one — it makes the analysis far sharper.

Start it: /next-best-action — or "what is the single best thing I could do for these people?"

What it does: analyses one household and names one action — unused tax space, a closing conversion window, RMD and IRMAA sequencing, Social Security and survivor math, an irreversible pension election, a protection or estate gap. Deliberately not your own open tasks read back to you.

What you get: the recommendation, the reasoning with every fact linked to its meeting, what would change the answer, who outside your firm has to act, and the three runner-ups with the one reason each lost.

Follow-ups: why this to expand the reasoning, what else for [client] to see it argue against itself.

Use it: when you know a household needs attention and want to decide what the next meeting is actually for.

Start it: /client-behavioral-profile — or "how should I approach these people?"

What it does: two depths on the same data, and it asks which you want. temperature [household] reads whether a relationship is cooling, steady or warming, and works from a single meeting. build_profile [household] is the full picture: motivators, fears, how and how fast they decide, who actually decides, whether stated risk tolerance matches behaviour, and engagement over time.

What you get: the read, plus a stance for the next conversation — the tone to take, an opening line in their own language, what to bring, and what to handle carefully. Every conclusion expands to the meeting behind it.

Use it: on the way into a meeting, or when handing a relationship to a colleague. Both outputs are internal — written for you, not for the client.

Start it: /client-journey — or "tell me everything about the Reynolds"

What it does: assembles everything on record about one family into a single view, polished enough to walk the client through it.

What you get: a five-stage lifecycle timeline, a relationship tree including beneficiaries you have never met, their outside professional team, a dated timeline of life events, and every motivator, worry and consequential fact with its corroboration and its age — plus what this household has never been asked that the rest of your book gets asked routinely.

Follow-ups: what's new on [client] for the short version.

Use it: in a review meeting, or to bring a new advisor up to speed on a family in ten minutes instead of a week.

Start it: /household-review-pack — or "prep the Thompsons' annual review"

What it does: builds the review from the last 12 to 24 months of meetings — goals then versus now, the decisions taken, every commitment you made with the date you said it, what changed in their life, and the handoffs the review creates for their CPA, attorney and other professionals.

What you get: an advisor summary covering all of it, each point linked to the meeting it came from.

Follow-ups: year_in_review for the client letter in a tone you choose, review_deck for a client-facing review presentation on your firm's own template, handoff for a brief to each outside professional.

Use it: a few days before an annual or semi-annual review. The letter and the deck are client-facing, so they are produced only when you ask for them.

Start it: /client-attrition-risk — or "who should I call this week?"

What it does: finds households that have gone silent past their own usual rhythm — a monthly client at 80 days is in more trouble than a semi-annual client at 100 — then ranks who to call first: anyone who named another advisor, left the last meeting concerned, was already sliding, or stopped bringing their own questions.

What you get: a ranked list, one card per household — why they are on it, how they left the room last time, who else is in the picture, what you said you would do and when, and a suggested opener in their own language.

Follow-ups: brief [client] for a full re-engagement brief, draft [client] for a re-engagement email you can copy. Nothing is ever sent for you.

Use it: Monday morning, as a call list. If nobody is past the threshold, it says so and stops rather than inventing urgency.

Questions, or a skill you would like to see? [email protected] or the in-app chat.