family-tree-mcp
This server lets you manage shared family-tree collections and publish them as a website.
List, create, rename, delete, and inspect trees
Add, update, merge, and remove people across trees
Manage tree membership without duplicating people
Add or remove parent, child, partner, and sibling links
Search people globally or within a tree
Review tree issues like duplicate names and placeholder parents
Calculate relationships, ancestors, and descendants
Undo the last collection-wide transaction
Export the full website locally, including all trees
Configure and publish the website, with living-person privacy options
Change the site-wide photo editing code and republish
Publishes an exported family tree site (a single read-only index.html) to a Cloudflare Pages project via wrangler, giving a free permanent https://.pages.dev link that can optionally be restricted with Cloudflare Access.
Generates Mermaid diagram source for the whole tree or a person's ancestors/descendants, and exports .mmd files with a mermaid.live link for downloading the chart as PNG/SVG/PDF.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@family-tree-mcp start a tree with my parents Maria and José, and grandma Rosa on my mom's side"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Family Tree
Self-hosted, private family trees with graph/list views, photos, search, relationship comparison, sharing and GEDCOM import/export. Public code; private family data.

Host your own
One application, two deployments. Follow setup & recovery.
Runtime | Node.js + SQLite + private files | Workers + D1 + private R2 |
Configuration |
|
|
Cost | Your hardware/hosting | Free within Cloudflare usage limits |
Visitors enter a shared family password. Approved editors and the configured owner sign in with Google. Access covers every tree in the installation; share links never bypass it.
Role | Permissions |
Viewer | Browse, suggest changes, add/replace photos |
Editor | Also edit people/relationships, review suggestions, clear photos |
Owner | Also manage trees/access/settings, import, delete people, export backups |
Related MCP server: Gramps MCP
Optional MCP
Requires Python 3.10+ and uv. Create a token under Agent access, then configure your MCP client to run uv run family-tree-mcp with:
Environment | Value |
| App URL including base path, e.g. |
| Owner-issued token, stored in client secrets |
Works with either host; edits save directly. Search before adding people and reuse their IDs across trees.
Develop
Node.js 26.12.1+, pnpm 12.15.0; Docker for production-image tests.
pnpm --dir web install --frozen-lockfile
pnpm --dir web build
pnpm --dir web devOpen http://localhost:8000/family-tree/; password: local-preview-only. The preview starts empty and stores data separately; screenshot demo data is not included.
Editing requires Google sign-in setup: GOOGLE_CLIENT_ID, OWNER_EMAIL and an authorized localhost origin.
Checks & code map · Operations · Architecture · MIT · ELK.js · Geist (OFL)
Available Tools
22 toolsadd_peopleAdd PeopleA
Add one or many people and their parent, partner and sibling links in one call.
Use a one-entry list to add a single person. Each entry can link to others in the batch by their ref (or name) and to existing people by id. All entries are added, or none are if any link is invalid.
Inputs: tree (required), people (required). Returns: object
change_photo_codeChange Photo CodeB
Replace the site-wide photo editing code and republish the complete website.
Inputs: code. Returns: object
check_treeCheck TreeARead-only
List things to review: children linked to only one of a couple, people with the same name, and placeholder parents. Report them to the user rather than fixing them unasked.
Inputs: tree (required). Returns: object
configure_websiteConfigure WebsiteA
Set website title and Cloudflare project independently of tree names.
Choose project before the first publication. Changing a published project requires a separate installation so its photo storage and URL cannot be reassigned by accident.
Inputs: name (required), project (required). Returns: object
create_treeCreate TreeB
Create a new, empty family tree.
Inputs: name (required), description. Returns: object
delete_treeDelete TreeADestructive
Delete only a tree definition. Shared people remain; undo restores the tree.
Inputs: tree (required). Returns: result
export_websiteExport WebsiteB
Build the complete site locally, including chooser and every tree, without deploying.
Inputs: hide_living. Returns: object
find_peopleFind PeopleARead-only
Find names matching every query word. Omit tree to search the shared collection.
Search globally before creating people; reuse existing UUIDs with set_tree_members.
Inputs: tree, query. Returns: result
get_ancestorsGet AncestorsARead-only
List a person's ancestors by generation (1 = parents, 2 = grandparents, ...).
Inputs: tree (required), person_id (required), generations. Returns: result
get_descendantsGet DescendantsARead-only
List a person's descendants by generation (1 = children, 2 = grandchildren, ...).
Inputs: tree (required), person_id (required), generations. Returns: result
get_personGet PersonARead-only
Get a person's details with their parents, partners, children and siblings.
Inputs: tree (required), person_id (required). Returns: object
get_relationshipGet RelationshipARead-only
Name how relative_id is related to person_id (e.g. "first cousin once removed").
Inputs: tree (required), person_id (required), relative_id (required). Returns: object
get_treeGet TreeARead-only
Return the whole tree: every person with their parent and partner ids.
Inputs: tree (required). Returns: object
get_websiteGet WebsiteBRead-only
Read site identity and publication status; never exposes the photo editing code.
Inputs: no named parameters. Returns: object
list_treesList TreesARead-only
List all trees with stable UUIDs, descriptions and membership counts.
Inputs: no named parameters. Returns: result
merge_peopleMerge PeopleADestructive
Merge a duplicate into keep_id, then remove the duplicate.
Applies across EVERY tree. Fills keep_id's missing details from the duplicate and moves its parents, partners and children. Also use it to replace a placeholder parent with the real person.
Inputs: tree (required), keep_id (required), duplicate_id (required). Returns: object
publish_sitePublish SiteA
Publish the ENTIRE configured website: chooser and all trees.
First call configure_website. All collection trees are public; living dates, places and notes are hidden by default. Automatic updates rebuild the entire site after edits pause. Publishing requires Cloudflare authentication; photos use a site-wide code.
Inputs: hide_living, keep_updated. Returns: object
remove_personRemove PersonADestructive
Delete a person and all relationships from EVERY tree.
To remove only their appearance in one tree, use set_tree_members(remove=True).
Inputs: tree (required), person_id (required). Returns: result
rename_treeRename TreeA
Rename a tree. People, backups and relationships are kept.
Inputs: tree (required), new_name (required). Returns: object
set_tree_membersSet Tree MembersA
Add existing collection people to a tree, or remove membership only.
Reuse UUIDs found with find_people (omit tree to search all people). Relationships are visible whenever both endpoints belong to the tree; no person is copied. Removing membership preserves the person's record and all global relationships.
Inputs: tree (required), person_ids (required), remove. Returns: object
undoUndoADestructive
Revert the last collection-wide transaction, including all affected trees.
This is global history, not per-tree history. Undone changes cannot be redone.
Inputs: no named parameters. Returns: object
update_peopleUpdate PeopleA
Change many existing people at once: details and renames, then links to add or remove.
Use for one or many person updates or relationship changes. Each link reads "relative_id is person_id's relation"; set remove=true to remove it. Everything applies, or nothing does if any entry fails, and undo reverts it all in one step. To add new people, use add_people.
Inputs: tree (required), links, changes. Returns: object
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 22 tools
Most tools target clearly distinct resources and actions, with explicit guidance separating overlapping areas such as remove_person vs. set_tree_members(remove=True), add_people vs. update_people, and publish_site vs. export_website. There are still several adjacent people/tree/website operations that an agent must parse carefully, but the descriptions generally resolve the boundaries.
Tool names follow a consistent verb_noun convention throughout: configure_website, create_tree, rename_tree, add_people, update_people, merge_people, find_people, get_relationship, export_website, and so on. The only minor deviation is the bare 'undo', but it is still a clear action name within an otherwise predictable pattern.
At 22 tools, the surface is heavy for a family-tree server, even though the tools cover several coherent subdomains: trees, people, relationships, website publishing, and review checks. Most tools earn their place, but the overall count sits in the borderline-heavy range and could likely be consolidated.
The core lifecycle is well covered: tree CRUD, people add/update/remove/merge/find/get, relationship queries, global undo, website configure/publish/export, and review checks. Minor gaps remain around backup/restore management, editing tree descriptions, and unpublishing an already published website, but these are not fatal to the main workflows.
Maintenance
Related MCP Connectors
Family history and genealogy in your AI: your tree and FamilySearch, with a source for every fact.
Read-only A2Me family context tools for AI assistants (members, dates, activity, relationships)
Access family personality profiles, relationship reports, and relational patterns via AI.
Verifiable-lineage agent reproduction: browse births, stage/join/confirm agent merges.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to create, edit, and query genealogical data from GEDCOM files. Supports complex genealogy searches, automatic data enrichment from web sources, relationship analysis, and biography generation for individuals and families.17MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to interact with Gramps genealogy databases for intelligent family tree research and management. Provides comprehensive tools for searching family data, creating records, analyzing relationships, and tracking genealogy research through natural language.52AGPL 3.0
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to query genealogy data from GEDCOM files with 24 tools for search, navigation, relationship analysis, and timeline events.1MIT
- AlicenseAqualityDmaintenanceReads and searches GEDCOM 5.5.1 files, enabling AI agents to look up individuals, families, sources, and multimedia references in a family tree.10MIT