Area (optional)
examples
Problem or use case
Some users learn from interactive notebooks (explore session JSONL, step through approve flow, visualize audit report). Flat .py scripts (Examples flat restructure) are better for CI and copy-paste; notebooks may help onboarding and docs — unclear if worth maintenance cost.
Proposed behavior
RFC / spike — decide yes/no before implementation:
| Option |
Pros |
Cons |
examples/notebooks/*.ipynb |
Interactive narrative, charts on export |
Extra dev dep, CI complexity, drift from SDK |
Notebooks in docs/ only |
Tied to docs sweep |
Not runnable as examples/ |
| No notebooks |
Simpler repo |
Less hand-holding for evaluators |
If yes: one notebook session_export_walkthrough.ipynb covering minimal loop + open .summary.json; optional [dev] extra jupyter; document in examples/README.md as supplementary (not replacing .py scripts).
If no: close with rationale in TESTING.md / CONTRIBUTING.
Alternatives considered (optional)
MyST or Quarto — heavier; defer.
Area (optional)
examples
Problem or use case
Some users learn from interactive notebooks (explore session JSONL, step through approve flow, visualize audit report). Flat
.pyscripts (Examples flat restructure) are better for CI and copy-paste; notebooks may help onboarding and docs — unclear if worth maintenance cost.Proposed behavior
RFC / spike — decide yes/no before implementation:
examples/notebooks/*.ipynbdocs/onlyexamples/If yes: one notebook
session_export_walkthrough.ipynbcovering minimal loop + open.summary.json; optional[dev]extrajupyter; document inexamples/README.mdas supplementary (not replacing.pyscripts).If no: close with rationale in TESTING.md / CONTRIBUTING.
Alternatives considered (optional)
MyST or Quarto — heavier; defer.