Stage 4 · Grades 11–12 · CS, capstone, independent study · in development
Build Your Research Agent
Computer science capstone on a public scholarly API
CANOA publishes a documented JSON API and machine-readable descriptions for AI agents. Students write a program — or an LLM-driven agent — that asks a real research question and defends its answer against the underlying data. The agent works with the open catalog only; no student data ever touches an AI system.
Project unit · 3–6 weeks
What students learn to do
- Read API documentation and paginate a real dataset responsibly.
- Write a script that aggregates thousands of records into one claim.
- Design a prompt or a tool loop that cites its evidence.
- Evaluate an AI answer against the source data and report where it fails.
Assignments in this track
| Task type | What the student produces |
|---|---|
| Curation project | Research question, method, results, and a data-backed critique. |
The material behind the track
128862 records available today. Every type is reachable through the public JSON API.
Standards alignment
- CSTA 3A-DA-11, 3B-DA-05 (create and refine computational models from data).
- AP Computer Science Principles: Create performance task (data and abstraction).
- ISTE Standards for Students 5 (Computational Thinker), 3 (Knowledge Constructor).
Alignments are our reading of the published frameworks, offered so you can start from something rather than nothing. Districts should map them against their own scope and sequence.