Setup and teaching materials for running the short course on the Hive HPC cluster, where each student gets a Hive account and drives a proteomics analysis through Claude Code.
https://bioshare.bioinformatics.ucdavis.edu/bioshare/view/short_course_2026/
Student Setup/ |
Everything students receive, plus the tooling that generates it |
final_handouts/1-share-with-everyone/ |
The current handouts, safe to post publicly |
hive/ |
Working copies of what's deployed to the shared course folder on Hive |
check-accounts.sh |
Verifies class accounts are ready (groups, Slurm association, home dir) |
provision.py |
Anthropic workspace create/archive (from the earlier per-workspace plan) |
instructor-setup-guide.md, HOW-TO-SET-UP (instructor).md |
The human guides |
hive/start-claude.sh is the launcher every student runs. It puts the whole Claude Code
session inside a Slurm allocation on a compute node before Claude starts, so no student
ever runs work on a shared login node.