ingest

Extracts entities, relationships, employment records, and observations from documents into a knowledge graph with provenance tracking.

Updated Apr 3, 2026
One-click install
npx skills add https://github.com/i-richardwang/Dossier --skill ingest-i-richardwang
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ingest
Source: https://github.com/i-richardwang/Dossier/tree/main/.claude/skills/ingest
Command: npx skills add https://github.com/i-richardwang/Dossier --skill ingest-i-richardwang

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

From articles, memos, and documents, this skill automatically extracts all factual data and upserts it into a knowledge graph, reducing manual curation and mistakes.

Core Features & Use Cases

  • Extracts entities, relationships, Employment records, and observations from unstructured sources.
  • Registers sources by URL with upsert semantics to preserve provenance and avoid duplicates.
  • Supports end-to-end workflows: ingest documents, enable research, reconcile conflicts, and analyze insights for projects.

Quick Start

Provide a document or URL and the project name as arguments to ingest, and the agent will extract all factual data and upsert it into the knowledge graph.

Frequently Asked Questions about ingest

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

FAQPage Schema
How do I extract entities and relationships from documents to populate a knowledge graph?

To extract entities and relationships for a knowledge graph, provide your unstructured documents and a project name to automatically upsert factual data, reducing manual curation and errors.

What is the best way to track provenance and avoid duplicate sources when ingesting research documents?

Tracking provenance and avoiding duplicates during document ingestion is handled through source registration by URL with upsert semantics, ensuring data integrity and preserving evidence timestamps.

Can I ingest employment histories and observations from multiple unstructured sources into a single knowledge graph?

Yes, ingesting employment histories and observations from multiple unstructured sources is supported, enabling end-to-end research workflows that reconcile conflicts and analyze project insights.

How do I register sources with upsert semantics for knowledge graph population?

Registering sources with upsert semantics requires providing the document URL and project name as arguments, allowing the agent to extract facts while preserving provenance and preventing duplicates.

Does knowledge graph ingestion from unstructured documents support evidence capture and timestamps?

Knowledge graph ingestion from unstructured documents fully supports evidence capture and timestamps, applying upsert semantics to maintain data integrity and track provenance across multiple sources.

What are the limitations of automatic entity extraction from unstructured research documents?

Automatic entity extraction from unstructured research documents requires source registration and upsert semantics to function properly, relying on accurate URLs and project names to prevent duplicate knowledge graph entries.