</> CAIS 2026 · DEMO PAPER
Skillful
Alhazen
A TypeDB-powered, ontology-backed notebook for agentic curation — one framework, many domains.
Brief
"The duty of the man who investigates the writings of scientists, if learning the truth is his goal, is to make himself an enemy of all that he reads, and, applying his mind to the core and margins of its content, attack it from every side."
— Ibn al-Haytham · 965–1039 AD
The Thesis
Scientists need structured notebooks, not just memory of conversations or plain-text notes.
— What most agent harnesses do
Save context to MD files or a RAG index. Adequate for chat, but breaks down when you need structured questions or tracking complex context at scale.
— What Skillful Alhazen does
Use Claude Skills to store agentic context to a next-generation ontologically-powered knowledge graph (TypeDB).
— The lesson, recalled
Bioinformatics taught us data hygiene 20 years ago. This was the justification for biomedical ontologies — the same issues apply now.
System Architecture
The Notebook Model
A simple, general high-level schema for curation.
The Alhazen notebook schema is a type hierarchy. Everything inherits from alh-identifiable-entity. Each skill extends this basic model with its own typed namespace.
Curation Pipeline
Three Hard Problems
Hard Problem · One
Use ontological distinctions to create clarity.
Most schemas conflate person with role. We apply BFO/UFO patterns: a person is rigid and intrinsic; a role is anti-rigid and externally grounded. Strip the institution — the role evaporates, the person remains.
Hard Problem · Two
Evolve domain models with experience.
Most agent stacks treat "I can't represent this" as a silent skip. We treat it as the most important signal.
- Execution Failure Crash, timeout, empty result.
- Schema Gap Claude detects missing concepts → filed as GitHub issue.
- User Request User requests schema or code changes.
Gaps are signal — how the graph grows from use.
Hard Problem · Three
Map memory data between schemas.
Claude authors GLAV mapping rules (Fagin, Kolaitis, Miller, Popa 2005) for two cases:
A · External Integration
Map complex data from external schemas into the notebook — e.g. Monarch's DisMech via TypeDB staging.
B · Schema Migration
Migrate notebook data between schema versions as the model evolves.
Deployed Skills
| Skill | Type | Domain | What it does |
|---|---|---|---|
| tech-recon | core | technology reconnaissance | Goal-driven investigation over code repos, papers, docs, web-pages |
| curation-skill-builder | core | domain modeling | Create schema, scripts, and dashboard for a new curation skill |
| jobhunt | external | career tracking | Track positions, fit analysis, skill-gap identification |
| coach | external | health tracking | Unify Apple Watch, lab reports, doctor notes into one schema |
| dismech | external | disease mechanisms | Ingest Monarch Initiative's DisMech (1000+ curated mechanisms) |
| scilit | external | scientific literature | Functionality for scientific publications (used by other skills) |
In Practice
Agentic 'memory' rendered from the knowledge graph.
Disease mechanisms, scientific literature, competitive intelligence, career tracking — one schema-validated store.
Outcomes
