init-project

Initializes GPS-conformant genealogy research projects from FamilySearch person data.

3|3|Updated Apr 8, 2026
One-click install
npx skills add https://github.com/PioneerAIAcademy/cowork-genealogy --skill init-project-pioneeraiacademy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: init-project
Source: https://github.com/PioneerAIAcademy/cowork-genealogy/tree/main/packages/engine/plugin/skills/init-project
Command: npx skills add https://github.com/PioneerAIAcademy/cowork-genealogy --skill init-project-pioneeraiacademy

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Starting a genealogy research project requires setting up structured audit-trail files and importing existing tree data without losing provenance. This Skill creates a complete, standards-compliant project workspace from a FamilySearch person ID or name search, so researchers begin with organized data instead of scattered notes. ## Core Features & Use Cases - Project scaffolding: Creates research.json (GPS audit trail with questions, plans, search log, assertions) and tree.gedcomx.json (simplified GedcomX deliverable) in one validated operation. - FamilySearch import: Fetches a person with relatives and source descriptions via person_read, re-ids everyone to local GedcomX identifiers, and attaches quality-rated source references to every fact and relationship. - Known-holdings survey: Records documents, prior research, photos, and oral knowledge the researcher already holds, preventing redundant searches later. - Pedigree analysis: Runs gap detection, a closed list of obvious-error checks, and a check-warnings pass, then reports findings in plain language. - Use Case: A user says "Start a new research project for person KWCJ-RN4. I want to identify his parents." The Skill fetches the tree data, builds the project files, surveys holdings, and delivers a summary with the first research gaps. ## Quick Start Start a new genealogy research project for FamilySearch person KWCJ-RN4 with the objective of identifying his parents.

Frequently Asked Questions about init-project

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I start a new genealogy research project from FamilySearch?▼

Provide a FamilySearch person ID or a name with identifying details, and the Skill calls person_read with relatives and source descriptions enabled, builds a simplified GedcomX tree, and creates research.json and tree.gedcomx.json via project_create in a single pass.

What if I don't have a FamilySearch person ID?▼

The Skill searches the Family Tree by name using person_search with a surname plus at least one other field such as given name, birth place, or a relative's name. If no candidates match, it initializes the project from your stated objective using local stub persons.

Can I use this on a folder that already has a research.json file?▼

No. A guard clause detects an existing research.json and stops before any file reads or tool calls, directing you to project-status or question-selection instead. The Skill never overwrites an existing project.

Why does the imported FamilySearch data get quality 1 sourcing?▼

FamilySearch Family Tree data is collaborative and unverified, so every imported fact and relationship receives a source reference with quality 1 (questionable). This marks the data as a starting point requiring verification, not an established conclusion.

What are the limitations of the initialized project?▼

Version 1 is read-only with respect to FamilySearch; the tree.gedcomx.json file is never uploaded back. Stub persons are created only for people the user named or directly implied, and unknown-gender stubs must include empty given names and Unknown gender to pass validation.