# Wormhole Analog Lab Notebook

This notebook records what we run inside the wormhole sandbox. Entries describe simulations and diagnostics only—they are not physical claims.

## Orientation
- `wormhole_smooth.py` is the primary simulator.
- Diagnostics live in `scripts/` and emit JSON/markdown under `artifacts/`.
- Visual dashboards reside in `wormhole_proof/wireframe/` and can serve the same artefacts via FastAPI.

## Working Guidelines
1. Prefer state-dependent geometry (local Jacobians, pullbacks) to fixed matrices.
2. Calibrate shared constants (κ, window radius) once per run and log them with the outputs.
3. Link every summary to the artefact that produced it; avoid unverifiable notes.
4. Flag QA traps (downsampling, change-point thresholds, viewer assumptions) when they appear.

## Current Focus
- Maintain the GR analogue scripts (`gr_metric_analysis.py`, `gr_hotspot_report.py`).
- Keep hotspot overlays in sync with the FastAPI diagnostics endpoint.
- Periodically re-run seeds {7,11,13,17,19} for regression checks and calibration validation.

## Recent Runs (2025‑11‑07)
- `python scripts/run_gr_suite.py --output artifacts/gr_suite --seeds 7 11 13 17 19 --overwrite`
- `python scripts/gr_metric_analysis.py --feed artifacts/smooth_gemini/wireframe.json --window 9 --kappa 5.20208 --out artifacts/relativistic_suite/gr_metric_seed7.json`
- `python scripts/pack_gr_metrics.py --base artifacts/relativistic_suite --feeds artifacts/gr_suite`
- `python scripts/gr_hotspot_report.py`
- `python scripts/gr_window_sweep.py --windows 5 9 13 --seeds 7 11 13 17 19`

Outputs are checked into `artifacts/relativistic_suite/` and `artifacts/relativistic_suite_ablation/` for post-hoc inspection.