What's Changed
- Claude code integration guide by @shreyashankar in #471
- optimizer: add directives for resolve operator by @shreyashankar in #470
- Fast Decomposition for Map Operations in DocWrangler by @shreyashankar in #472
- Embedding blocking threshold optimization by @shreyashankar in #473
- fix broken links in python api documentation by @shreyashankar in #476
- feat: add token usage to pipeline runner by @shreyashankar in #480
- fix: pin pyrate-limiter<4 to prevent breaking import changes by @shreyashankar in #483
- Web fetch and search by @redhog in #486
- Interactive progress view for pipeline runs (terminal) by @shreyashankar in #488
- Simplify MOAR interface and add Python API by @shreyashankar in #490
- Model cascades: cut LLM cost on binary operators with statistical quality guarantees by @shreyashankar in #491
- Remove combined precision+recall guarantee by @shreyashankar in #494
- Introduce PySpark-like Frame API and refactor runner architecture by @shreyashankar in #495
- Agentic LLM tool support by @shreyashankar in #499
- chore: bump version to 0.3.0 for PyPI release by @shreyashankar in #501
Full Changelog: 0.2.6...0.3.0