RootsMagic 11 AI prompts for cleaning imported genealogy sources

You can use the AI Prompt Builder plus an external AI to standardize, de-duplicate, and reformat messy imported sources in RootsMagic 11. Below are ready-to-use prompt patterns you can copy and paste into the Prompt Builder and reuse.rootsmagic+1

How to set these up inside RootsMagic 11

• Open any person, then open the AI Prompt Builder from the AI menu or person tools.rootsmagic+1

• Choose to create a free-form prompt, then paste one of the prompt texts below into the editor.[help.rootsmagic]​

• In the options, decide whether to include that person’s events/relatives or leave the prompt generic; for pure source cleanup, you usually do not need person data included.rootsmagic+1

• Save with a name like “Clean imported source,” “Normalize online tree citation,” or “Source template extractor,” then reuse it for any person or fact by selecting the prompt and clicking Copy.rootsmagic+1 

1. General “clean this source” prompt

Use this when you have a poorly formatted citation or source text pasted into a note or fact.

Prompt text (save as a reusable prompt):

Role: You are a professional genealogist and citation editor specializing in Evidence Explained–style citations. You will receive messy, imported source or citation text from RootsMagic.

Task:

1. Identify and correct obvious typos and spacing.

2. Standardize capitalization (title case for titles, sentence case for notes).

3. Expand abbreviations only when the meaning is clear (e.g., “Co.” → “County” when used in a place).

4. Separate the information into:

   • Source list entry (for RootsMagic “Master source” field).

   • Full reference note (first citation).

   • Subsequent reference note (short citation).

5. Do not invent missing details. If something is missing or unclear, write “[unknown]” or “[illegible]” rather than guessing.

Output format (use this exact structure):
Source list entry:
[one line]

Full reference note:
[one paragraph]

Subsequent reference note:
[one line]

Here is the messy source text to clean and structure:

text
<<<PASTE ROOTSMAGIC SOURCE OR CITATION TEXT HERE>>>  

2. Prompt to normalize imported online trees / Ancestry sources

Use this when imports have long, redundant “Ancestry Family Tree” style citations.

Role: You are a professional genealogist. Your task is to normalize messy combined-source citations imported from online trees into RootsMagic.

Task:

1. Remove marketing or irrelevant phrases (e.g., “Ancestry Family Trees,” “online publication,” “Provo, Utah, USA” publisher boilerplate) unless needed for citation clarity.

2. Identify the underlying record type (census, vital record, city directory, etc.) and rephrase the citation so the record type is clear.

3. Extract and clearly label: database/collection name, jurisdiction, record type, year or range, and access URL (if present).

4. Create:

   • A cleaned “Master Source” text suitable for RootsMagic (collection-level).

   • A “Citation detail” template line showing how to plug in person-specific details (e.g., “line ___, dwelling ___, family ___”).

5. Do not invent missing fields; mark them with “[unknown]”.

Output format:
Master source:
[one line, collection-level]

Citation detail pattern:
[one line showing structure for the specific citation]

Cleaned example citation (for this person only):
[one paragraph]

Here is the imported source text:

text
<<<PASTE IMPORTED ANCESTRY / OTHER TREE SOURCE TEXT HERE>>>  

3. Prompt to extract data into a RootsMagic source template

Adapt this pattern to whichever built‑in template you are using (similar to the “Artifact, privately held” and “Newspaper, unidentified clipping” extraction prompts people are using with RM11).youtube+1

Role: You are a professional genealogist and forensic document analyzer. Your task is to extract structured data from the text that follows so that I can populate a specific RootsMagic 11 source template.

Template type: [e.g., “Census, U.S. Federal (online images)” or “Artifact, privately held”].

Task:

1. Extract values for these fields (use the exact labels and order):

   • Source name

   • Author / creator

   • Title or description

   • Jurisdiction (state, county, place)

   • Repository or website

   • Call number / URL / path (if any)

   • Item type (e.g., deed, certificate, clipping, artifact)

   • Date of record or event

   • Person(s) of interest

2. If a field is not present, output “[not stated]”.

3. Do not add facts that are not explicitly supported by the text.

Output format (exact headings):
Source name:
Author / creator:
Title / description:
Jurisdiction:
Repository / website:
Call number / URL / path:
Item type:
Date of record or event:
Person(s) of interest:
Notes: [any clarifying comments or uncertainties]

Here is the text to analyze:

text
<<<PASTE IMPORTED SOURCE OR CITATION TEXT HERE>>>  

Once you have this output, you can copy–paste each field into the corresponding RM11 template fields.[youtube]​[help.rootsmagic]​

4. Prompt to de-duplicate and consolidate similar sources

Use this when imports have multiple near-identical master sources.

Role: You are a professional genealogist and data curator. Your task is to analyze multiple RootsMagic source descriptions that may represent the same underlying record collection.

Task:

1. Group the following source descriptions into clusters that refer to the same collection.

2. For each cluster, propose a single standardized “Master Source” label suitable for RootsMagic.

3. List which original descriptions belong to each cluster.

4. Do not guess; if you are unsure whether two items are the same, put them in separate clusters and mark them “possibly same collection.”

Output format:
Cluster 1 – Master Source label:
[proposed standard label]
Includes:

• [original source text 1]

• [original source text 2]

Cluster 2 – Master Source label:
[etc.]

Here are the imported source descriptions (each separated by a blank line):

text
<<<PASTE A LIST OF SOURCE TITLES / DESCRIPTIONS HERE>>>  

You can then manually merge or retitle master sources in RM11 based on the clusters.[help.rootsmagic]

__________________________________

Here’s a tailored AI prompt you can use with the RootsMagic 11 “Website (with multiple databases)” template to clean imported sources and extract structured fields. You can save this as a reusable prompt in the AI Prompt Builder.[youtube]​[sqlitetoolsforrootsmagic]​

Prompt for “Website (with multiple databases)” template

Role: You are a professional genealogist and citation editor working to Evidence Explained standards. You will receive messy text from an online collection (e.g., Ancestry, FamilySearch, MyHeritage, Fold3) that should be cited in RootsMagic 11 using the “Website (with multiple databases)” source template.

Task:

1. Identify the specific database/collection on the site and treat it as the main source.

2. Clean the text (fix typos, spacing, capitalization) without changing meaning or adding facts.

3. Extract values for the RootsMagic “Website (with multiple databases)” template fields listed below.

4. If a field is not present or is unclear, output “[not stated]” and do not guess.

Fields to extract (Master Source level – the database):

• Source name (recommended pattern: “[Year range] [Database title] – [Website title]”)

• Database author / creator (organization or person)

• Database title (collection name as given on the site)

• Creation or owner (entity that created or owns the digital collection, if given)

• Title of website (e.g., Ancestry, FamilySearch, MyHeritage, Fold3)

• URL for database home page (stable collection-level link, not the person page if possible)

• Item type (e.g., “database with images,” “database,” “digital images”)

• Credit line (archive, microfilm provider, or original record holder, if stated)

• Source text (brief description of what the database contains, including place and time span, based only on the provided text)

• Source comment (notes on coverage, quality, or indexing caveats, if present)

• Source reference number (any catalog number, film number, DGS number, or internal ID, if given)

Fields to extract (Citation / item of interest level – this person’s record):

• Access type (e.g., “digital image,” “indexed database with images,” “indexed database”)

• Access date (in form “Day Month Year,” e.g., “22 March 2026,” if present)

• Item of interest (name of person of interest and key identifiers such as event type, date, and place, based only on the supplied text)

• Website path or record URL (direct link to the individual record page, if given)

• Notes about the specific item (any remarks specific to this person’s record, such as “image truncated,” “index only,” or “original language Spanish,” if stated)

Output format (use these exact headings and keep them in this order):
Master source – Website (with multiple databases)
Source name:
Database author / creator:
Database title:
Creation or owner:
Title of website:
URL (database home page):
Item type:
Credit line:
Source text:
Source comment:
Source reference number:

Citation – Item of interest
Access type:
Access date:
Item of interest:
Record URL or path:
Item notes:

Here is the text to analyze and clean (this may include an imported RootsMagic citation, collection description, and/or record page details):

text
<<<PASTE IMPORTED ONLINE SOURCE / CITATION TEXT HERE>>>  

You can then copy each output line into the corresponding fields in the “Website (with multiple databases)” template in RootsMagic 11, using one Master Source per collection and separate citations for each person’s record.wiki.rootsmagic+1

___________________________________

 Here are three RM11-ready AI prompts for U.S. census citations

This one is tuned for the kind of work you’re doing. You can adapt wording for 1850–1950 and online vs microfilm.

 Role: You are a professional genealogist and citation editor working to Evidence Explained standards. You will receive messy text describing a U.S. census record (often copied from an online tree or site). The record will be cited in RootsMagic 11 using an online‑database census template.

Task:
Identify the census year, country, state, county, and locality; identify whether the site provides images, an index only, or both.

Clean the text (fix typos, spacing, capitalization) without changing meaning or adding facts not stated in the text.

Treat the database or collection as the Master Source and the individual household/entry as the Citation.

Extract values for these fields. If a field is not present, output “[not stated]” and do not guess.

Master Source (collection-level – for RootsMagic “U.S. census, online images” template):

Source name (recommended: “[Year] U.S. census, [state] – [website]”)

Census year

Country

State or territory

Database / collection title (as given on the site)

Website title (Ancestry, FamilySearch, etc.)

Publisher / host (if given)

Database description (what the collection covers, including years and record type, based only on the provided text)

Credit line or original record creator (e.g., “National Archives and Records Administration [NARA], microfilm publication ___,” if stated)

Database‑level URL (home page or search page for this census collection)

Notes about coverage or limitations (e.g., missing counties, index only), if stated

Citation (household or person-level):

Person(s) of interest (focus person’s name and role in household)

Jurisdiction (state, county, town or township as given)

Enumeration details (e.g., enumeration district, ward, city, etc., if given)

Page and location information (page, sheet, line, dwelling, family numbers, as provided)

Schedule type (e.g., population schedule)

Event date or census date (as stated in the collection or text)

Record identifier(s) (e.g., NARA microfilm publication and roll number, image number, digital film number, or other site‑specific ID)

Access type (image, index, or both)

Access date (e.g., “22 March 2026,” if given, otherwise “[not stated]”)

Direct record URL (if any)

Notes about legibility, indexing errors, or variant spellings, if mentioned in the text

Output format (use these exact headings and keep this order):
Master source – U.S. census (online images)
Source name:
Census year:
Country:
State or territory:
Database / collection title:
Website title:
Publisher / host:
Database description:
Credit line / original creator:
Database URL:
Database notes:
Citation – household / person
Persons of interest:
Jurisdiction:
Enumeration details:
Page / sheet / line / dwelling / family:
Schedule type:
Census date / event date:
Record identifiers (film / roll / image / DGS / etc.):
Access type:
Access date:
Record URL:
Item notes:
Here is the imported census source / citation text to clean and analyze:

<<<PASTE IMPORTED CENSUS SOURCE OR CITATION TEXT HERE>>>​