hex-ob — HEX Ophyd-async + Bluesky for Dev

Workspace for developing and testing hextools and hex-profile-collection against the simulated HEX beamline (hxm_program/simulated_beamlines/HEX) without touching the production repos (Nghia's setup, 2026-07-30). Break things freely here; merge back when green.

Layout

hex-ob/
  hextools/                 # full clone (own .git, all branches, GitHub remotes)
  hex-profile-collection/   # full clone (own .git, all branches, GitHub remotes)
  README.md                 # this file — sandbox-level docs live at this level

The two package directories are independent nested git repositories, deliberately ignored by hex-ob's own git (see .gitignore). hex-ob tracks only sandbox-level documentation/glue; the packages keep their own history so the path back to production stays clean.

The contract

1. Develop here. Work on branches inside hextools/ / hex-profile-collection/ (current dev branches: hxm1288-f1-common-scaffold, hxm-1288-sim-boot). Guided-tutorial work for HXM-1288 tomography (AJ implements, agent reviews) happens in these copies.
2. Test against the sim. The simulated beamline (PandA Tomo_radio_1_config, real AD IOC frame tier, motor + bridge) is the test bed; the pyepics tomo_flyscan.py oracle outputs are the functional-equivalence reference (DECISIONS.md D-0014).
3. Merge back via PRs. When something is production-ready, push its branch to the fork remote and open a PR against the NSLS2 repo (AJ authors, Nghia reviews/merges — D-0013). Nothing lands in production by side effect.

Environments

Pixi/venv environments were deliberately NOT copied (they embed absolute paths). Recreate in place:

cd hex-profile-collection && pixi install     # profile envs
cd hextools && uv sync                        # hextools (develop against uv.lock)