v0.4.13

## What's Changed
* feat(lang): add opt-in detection reasons and source diagnostics by @xinpingzh in https://github.com/NandhaKishorM/laya/pull/1284
* fix(hooks): preserve label text when renaming options without descriptions (#1285) by @Swatantra-66 in https://github.com/NandhaKishorM/laya/pull/1286
* fix(hooks): keep stability calibration and replay on the same precision by @xinpingzh in https://github.com/NandhaKishorM/laya/pull/1289
* feat(ts): add local option stability diagnostics by @xinpingzh in https://github.com/NandhaKishorM/laya/pull/1290
* fix(dotnet): match the Python SDK exactly in routing, language, email, rounding and config by @kevin-gatimu in https://github.com/NandhaKishorM/laya/pull/1291
* feat(rust): add predict_long and match the Python SDK exactly by @kevin-gatimu in https://github.com/NandhaKishorM/laya/pull/1292
* fix(tools): fail closed on a non-object remote decision response by @aashish254 in https://github.com/NandhaKishorM/laya/pull/1293
* fix(structured): refuse a boolean score bound instead of reading it as 1/0 by @aashish254 in https://github.com/NandhaKishorM/laya/pull/1294
* fix(evals): refuse a boolean confidence in _answer_confidence by @aashish254 in https://github.com/NandhaKishorM/laya/pull/1295
* fix(agno): reject a non-object remote decision response by @aashish254 in https://github.com/NandhaKishorM/laya/pull/1296
* feat(sdk): add Python HTTP batch predictions by @xinpingzh in https://github.com/NandhaKishorM/laya/pull/1297
* perf(agent): optimize logit decoding and confidence calculation in _decode_answers by @HamidRezaLS in https://github.com/NandhaKishorM/laya/pull/1299


**Full Changelog**: https://github.com/NandhaKishorM/laya/compare/v0.4.12...v0.4.13