← Skillful-Alhazen

TypeDB Notebook

Skills: TypeDB Notebook

The /typedb-notebook skill provides core knowledge operations for the Alhazen system. It’s the foundational skill that other skills build upon.

Overview

This skill handles the fundamental operations:

  • Remember - Store findings with provenance
  • Recall - Query what you know about topics
  • Organize - Manage collections and relationships

Quick Start

Remember a Finding

You: Remember that the Smith et al. paper found lipid nanoparticles
     were 3x more effective than viral vectors for liver targeting

Agent: [Creates note with provenance]

Noted. I've stored this finding and linked it to the Smith et al. paper.
The note includes:
- Content: Your finding about 3x effectiveness
- Source: Smith et al. paper
- Timestamp: Now
- Type: Research finding

Recall Knowledge

You: What do I know about lipid nanoparticles?

Agent: [Queries notes, papers, fragments]

From your knowledge graph:

**Notes:**
- LNPs are 3x more effective than viral vectors for liver (Smith et al.)
- Ionizable lipids are key component for endosomal escape (Chen et al.)

**Papers in collections:**
- 5 papers in "CRISPR Delivery Review" mention LNPs

**Related topics:**
- mRNA delivery
- Gene therapy
- Liver targeting

Organize Collections

You: Create a collection called "PhD Literature Review"

Agent: [Creates collection]

Created collection "PhD Literature Review".

You: Add the paper with DOI 10.1038/s41586-023-00001-1 to it

Agent: [Adds paper to collection]

Added to "PhD Literature Review".

Note Types

You can create different types of notes:

Type Purpose Example
finding Key result or claim “LNPs are 3x more effective…”
question Open research question “Why do LNPs work better in liver?”
critique Critical analysis “Small sample size in Table 2…”
connection Link between ideas “This relates to Chen’s work on…”
summary Synthesis of sources “Three papers agree that…”
todo Action item “Read the methods section”

Core Entities

Collections

Named groups for organizing knowledge:

You: What collections do I have?

Agent: [Lists collections with counts]

Your collections:
- PhD Literature Review (23 papers, 12 notes)
- CRISPR Delivery (5 papers, 8 notes)
- Job Hunt Q1 (handled by /jobhunt skill)

Things

Any item you’re tracking:

  • Papers
  • Datasets
  • Tools
  • People
  • Organizations

Artifacts

Raw captured content with provenance:

  • PDF documents
  • Web page snapshots
  • API responses
  • Email threads

Fragments

Extracted portions of artifacts:

  • Quotes
  • Figures
  • Tables
  • Key sentences

Notes

Your annotations about any entity:

  • Findings
  • Questions
  • Critiques
  • Summaries

Provenance Chain

Every piece of knowledge traces back to its source:

Artifact (PDF of Smith et al. paper)
    ↓
Fragment (quote from Results section)
    ↓
Note (your finding about 3x effectiveness)

You can always ask: “Where did this come from?”

You: Where did the 3x effectiveness finding come from?

Agent: [Traces provenance]

This finding comes from:
- Your note created on 2024-02-01
- Based on fragment from Smith et al. Results section (page 7)
- Original artifact: PDF downloaded from Nature.com

Search and Query

Text Search

You: Search my notes for "effectiveness"

Agent: [Full-text search across notes]

Found 3 notes mentioning "effectiveness":
1. LNPs 3x more effective for liver... (Smith et al.)
2. AAV shows higher effectiveness in CNS... (Jones et al.)
3. Comparison of effectiveness across tissues... (your synthesis)

By Entity Type

You: List all papers I have notes about

Agent: [Queries papers with linked notes]

Papers with your notes:
- Smith et al. 2023 - 4 notes
- Jones et al. 2022 - 2 notes
- Chen et al. 2024 - 1 note

By Time

You: What did I add last week?

Agent: [Queries by timestamp]

Last week you added:
- 3 papers to "CRISPR Delivery"
- 5 notes about delivery mechanisms
- 1 new collection "Vaccine Development"

Integration with Other Skills

The typedb-notebook skill underlies all domain skills. Every skill that stores data in TypeDB uses the same identifiable-entity hierarchy and core relations:

Think of typedb-notebook as the operating system for your knowledge graph — it provides the core remember, recall, and organize operations that work uniformly across all domain types.