To request access, please have the admin user contact support@precisefp.com.

Ask is the simplest way to use PreciseFP MCP: type a question in Claude or ChatGPT and get an answer from your live PreciseFP data. No report builder, no export, no support ticket.

If you haven’t connected yet, start with How to activate PreciseFP MCP.

Two questions worth asking on day one:

Who am I in PreciseFP?

Returns your name, role, team, and contact details.

What can you do in PreciseFP?

Returns a summary of what it can reach — accounts, forms and engagements, activities, pipelines, and users.

How many accounts do I have?

Find the account for Margaret Chen.

Show me my prospects.

Which accounts are in my Discover stage?

Account lookups are fuzzy and forgiving — a misspelled name or a partial email will usually still find the right record.

This is where most advisors get immediate value, because engagement status is normally several clicks away.

What engagements are currently in progress?

Which engagements haven’t been started yet?

How complete is the data on my in-progress engagements?

Has the Chen household completed their risk tolerance form?

You’ll get the engagement, the recipient, its status, how much has been filled in, and a live access link if you want to open it.

One thing this surfaces well: engagements that never started because the email didn’t reach the client. Asking “which engagements have not been started?” is a fast way to catch delivery problems you’d otherwise never notice.

What has my team been doing today?

What has Sarah done in the last month?

Show me all form sends from last week.

You’ll get a readable activity history — logins, form sends, template edits, profile updates, integration events — with who did it and when.

What pipelines do I have configured?

What are the stages in my client pipeline?

What triggers move an account from Onboard to Discover?

Which integrations do I have enabled, and when were they turned on?

Pipeline questions are useful for auditing automation you set up months ago and haven’t looked at since.

Which tools answer which questions, and what their limits are.

Your identity and your team. “Who am I?” resolves through users-current, which returns your identity, role, and team. users-list enumerates everyone you can see, with pagination.

Activity history. Two separate audit trails back the activity questions:

Both are append-only and both record the channel of each change — mcp, api, engagement, or integration — so you can tell whether a field was edited by hand, imported, filled in by a client, or written by an integration. They also record impersonated sessions, so support activity on your account is visible to you.

Because there is no dedicated integrations tool, users-activities is the authoritative source for reconstructing which integrations were enabled and when. That’s why the integrations question above is answered from activity history rather than a settings list.

Account lookups. accounts-search does the fuzzy, typo-tolerant, relevance-ranked matching. accounts-list enumerates with pagination and filters for CLIENT, PROSPECT, FAVORITES, or ARCHIVED. accounts-get fetches one account by ID.

Important limitation: these three return metadata only — names, contact details, a data-quality score, pipeline placement, and timestamps. They deliberately do not return financial data. Birthdates, addresses, income, assets, liabilities, and insurance all sit behind a separate profile lookup, which is covered in the Analyze article.

Engagements. form-engagements-list filters by NOT_STARTED, IN_PROGRESS, or COMPLETED. form-engagements-search matches on contact name, email, or template title. Each engagement carries a filled count — the number of fields the recipient has entered. A nuance worth knowing: a partially filled engagement doesn’t flow into the account profile until the client submits it. An engagement can sit in progress with data entered for months while the account profile shows none of it. Engagement filled and account quality score are independent signals — don’t read one as the other.

Pipelines. dataset-pipelines returns stage definitions for the CLIENT and PROSPECT tracks, and in detail mode the automation rules behind them. A rule pairs a trigger — an engagement of a given title completing, or an account sitting in a stage for N days — with an action, either advancing a stage or converting a prospect to a client.

Why does it sometimes give a different list of capabilities? The capability summary is generated by the AI model from the tool descriptions rather than being a fixed list, so the wording varies between models and tools. The underlying capabilities are the same.

Can it see other advisors’ clients? No. Every call runs as you, inside your organization, using your existing PreciseFP permissions.

Does asking questions change anything? No. Everything in this article is read-only.

Can I trust the numbers? The data is live from PreciseFP. Just be careful with aggregate figures across accounts that don’t all have profiles loaded — see Analyze for why that matters and how to ask for it properly.