Ready-to-use, Probate‑Specific Pompts for Gemini,

Here are ready-to-use, probate‑specific prompts for Gemini, organized by task. You can copy‑paste and adapt them for different estates, localities, or record sets.

General setup prompts (before you start)

Use one of these to “frame” the session so Gemini understands your genealogical goals:

<prompt>

You are assisting with documentary genealogy research using probate records.

I will upload or paste wills, inventories, accounts, and related court papers.

Your job is to:

1) transcribe or summarize the documents faithfully;
2) identify people, relationships, places, property, and dates;
3) highlight anything relevant to family structure, migration, or land;
4) point out ambiguities or conflicts I should verify in the original images.

Do not invent facts or laws. Always separate what the document actually states from your inferences.

</prompt>

- Locality‑aware frame

<prompt>

Act as a careful research assistant for a genealogist working in [county, state/country] probate records, [approximate date span].

When I give you a document, interpret it in plain language and explain any terms that might have a special meaning in probate practice of that era (e.g., relict, reversion, dower, curtesy).

Always keep track of names, roles, relationships, dates, and places for later correlation.

</prompt>

___________________________________________________________________________

Use one of these “frames” once at the start of a conversation, then move to the task‑specific prompts below.

Transcription prompts (handwriting and structure)

Gemini 3.x is particularly effective for handwriting transcription when you upload clear images instead of PDFs and give explicit rules. [denyseallen.substack](https://denyseallen.substack.com/p/practical-ai-toolkit-genealogy-2026-chatgpt-claude-gemini-perplexity)

- Exact‑spelling transcription

<prompt>

I am uploading one page of a handwritten probate record (likely a will) from about [year] in [jurisdiction].

Please create a line‑by‑line transcription with these rules:

– Preserve original spelling, capitalization, abbreviations, and punctuation.

– Do not modernize anything.

– If a word is unclear, write `[illegible]` or `[guess: word?]`.

– Keep line breaks roughly aligned with the original.

At the end, list any words or names you are uncertain about.

</prompt>

_____________________________________________________________________________

- Segmented transcription (for long packets)

<prompt>

I will upload several images from the same probate packet.

For each image, provide:

1) A short label for the document type (e.g., “Will, page 1”, “Inventory page 2”, “Final account”).
2) A faithful transcription (same rules as before: exact spelling, `[illegible]` for unclear words).
3) A one‑sentence summary of what happens on that page.

Keep each image’s output clearly separated with headings.

</prompt>

__________________________________________________________________________________

- Targeted clause transcription

<prompt>

From this probate image, transcribe only the clause(s) describing how the estate is to be distributed (who receives what).

Preserve the exact wording and spelling.

Then give me a plain‑language explanation of that clause.

</prompt>

________________________________________________________________________________

Entity, relationship, and role extraction

These prompts help you turn long probate text into structured data you can bring into RootsMagic, spreadsheets, or research logs.

- Person & relationship table

<prompt>

Here is the full text of a probate document (will, inventory, accounts) for [testator’s name], dated about [year].

Extract all named people and list them in a markdown table with columns:

`Name | Role in document (testator, executor, child, witness, creditor, etc.) | Stated or implied relationship to testator | Residence (if given) | Evidence snippet.`

Quote or closely paraphrase just enough text in the evidence column to show why you assigned that role or relationship.

If a relationship is inferred rather than explicitly stated, mark it “inferred”.

</prompt>

_______________________________________________________________________________

- Heirs and legatees focus

<prompt>

From this will and any related distribution papers for [testator’s name], identify all heirs and legatees.

Create a list with:

– Full name

– Relationship to testator (as written; say “not stated” if not explicit)

– What they received (e.g., land description, personal property, money)

– Whether the gift appears residuary, specific, or conditional.

After the list, briefly discuss what these gifts suggest about birth order, family structure, and economic status.

</prompt>

_____________________________________________________________________________

- FAN‑club extraction

<prompt>

Using this transcribed probate packet, identify the FAN‑club (Friends/Associates/Neighbors) for the testator.

List:

– Executors/administrators

– Witnesses

– Appraisers

– Creditors named

– Buyers at the estate sale (if present).

Put them into a table with: `Name | Role | Any stated residence or occupation | Document page reference.`

At the end, note 3–5 people who appear multiple times and might be worth additional research as potential kin or close associates.

</prompt>

____________________________________________________________________________

Property, land, and economic analysis

These prompts are useful when wills and inventories are key to reconstructing land holdings or economic context.

- Inventory classification

<prompt>

From this estate inventory, classify each item into categories such as: livestock, crops/food, tools, household goods, books, enslaved persons (if present), debts owed to the estate, and any other category that makes sense.

Output a table with: `Item description | Category | Quantity (if given) | Any value or price stated`.

Then give a 1–2 paragraph summary of what this inventory suggests about the family’s occupation(s), social status, and way of life.

</prompt>

_______________________________________________________________________________

- Land description extraction

<prompt>

From this will and any attached probate documents, pull out all land references (tracts, lots, watercourses, neighbors).

Create a list with:

– Exact wording of each land description

– Who is receiving or disposing of the land

– Any adjoining landowners, watercourses, or landmarks mentioned.

At the end, suggest how I might map these descriptions against modern maps or deed records.

</prompt>

_______________________________________________________________________________

- Debt and creditor profile

<prompt>

Using this probate file, identify all debts owed by the estate and all debts owed to the estate.

Create two tables:

Briefly comment on what these debts suggest about the testator’s business activities or social network.

</prompt>

______________________________________________________________________________

Chronology, correlation, and conflict checking

These prompts help Gemini do the tedious correlation across multiple documents inside a packet.

- Estate timeline

<prompt>

You will receive several related probate documents (will, bond, inventory, account, sale list, final settlement) for [testator’s name].

Build a timeline of the estate from the earliest to latest record, with entries like:

`Date | Event | Document/page | People involved | Brief note on genealogical relevance.`

If a date is only approximate (e.g., term of court), indicate that clearly.

- </prompt>

___________________________________________________________________________

Status of minor children

<prompt>

From these probate records, identify all references that suggest a child is a minor (e.g., guardianship appointments, age statements, schooling, apprenticeships).

For each child, list: `Name | Evidence for being a minor | Any explicit ages or “under 21” phrases | Implications for birth year range.`

Then propose a reasonable birth-year range for each minor child, clearly labeling it as an estimate derived from the documents.

</prompt>

___________________________________________________________________________

- Conflict and ambiguity review

<prompt>

Read through this compiled probate packet as a whole.

Identify any conflicts or ambiguities related to:

– Names (variant spellings, multiple individuals with same name)

– Relationships (e.g., “kinsman,” “cousin,” “friend” with unclear meaning)

– Property (discrepancies between will, inventory, and final distribution)

– Dates (contradictory or implausible date sequences).

Summarize each conflict in a short numbered list and suggest specific follow‑up records or checks I should perform to resolve them (e.g., look for guardianship dockets, land re‑surveys, or later suits in equity).

</prompt>

______________________________________________________________________________

Turning probate packets into research notes and to‑dos

Here are prompts that move from document content to practical genealogical work products.

- Research‑note summary

<prompt>

Based on this probate packet, draft a research note suitable for my genealogy file on [testator’s name].

Include sections:

– “Source summary” (what record set this is, jurisdiction, volume/page or microfilm/image citation; if missing, tell me what I still need to look up)

– “Key genealogical information” (people, relationships, dates, places)

– “Negative or missing information” (what we might expect but don’t see)

– “Next steps” (3–7 specific research actions based on these records).

Do not invent citation details; if unknown, leave placeholders like `[volume?]` or `[image number?]`.

</prompt>

__________________________________________________________________________

- Heir chart for proof argument

<prompt>

Using the heir and legatee information from this will and related papers, create a heir‑distribution chart that I can adapt into a proof argument.

Output a table:

`Heir name | Relationship (stated or inferred) | Evidence from probate | Share of estate (qualitative is fine: large/small/residuary) | Notes for analysis`.

Then briefly discuss how this pattern of distribution supports or challenges the hypothesis that [state your hypothesis].

</prompt>

__________________________________________________________________________

- To‑do list generation

<prompt>

From what you see in these probate records, generate a prioritized to‑do list of further records I should seek to solidify this family group.

Group tasks under headings like “Land records,” “Court records,” “Church or cemetery records,” and “Other,” and explain in one sentence per item why it might help.

</prompt>

______________________________________________________________________________

Working with images vs. text

To get the best results, combine upload and text prompts like this:

- Image upload plus immediate instruction

<prompt>

I am uploading a high‑resolution image of a handwritten probate will.

First, confirm whether the image quality is sufficient for a careful transcription.

If it is, proceed with a line‑by‑line transcription preserving original spelling and punctuation.

If it is not, explain what is hard to read and suggest ways I might enhance the image before trying again.

</prompt>

_____________________________________________________________________________

- Check Gemini’s work against your reading

<prompt>

Here is your earlier transcription of this probate image, followed by my own manual transcription.

Please compare them and list all differences in wording, spelling, or line breaks.

For each difference, indicate which reading is more plausible based on normal handwriting patterns and the surrounding context—but do not claim certainty.

</prompt>

______________________________________________________________________________

Using Flash‑Lite vs. Pro (for big batches)

When you have many pages, use a two‑step workflow with Flash‑Lite then a richer model. [gemini](https://gemini.google/explore/)

- Flash‑Lite for quick summaries

<prompt>

I will paste text from multiple probate documents.

For each one, give me a 1–2 sentence summary and identify: testator, document type, main people, and whether there is any explicit relationship statement (e.g., “my son,” “my daughter”).

Label each summary with “Doc 1,” “Doc 2,” etc., so I can match them back.

</prompt>

______________________________________________________________________________

- Pro/Thinking for deep analysis

<prompt>

Here are summaries and selected full transcriptions from a probate packet, already labeled by document.

Using deep analysis, answer:

1) What is the most likely nuclear family group for [testator’s name]?
2) Which individuals might be grandchildren or collateral relatives instead of children, and why?
3) What gaps remain in the evidence that I should address before writing a proof argument?

</prompt>

Search This Blog

PROMPT OF THE DAY