v1.6.39
ShenhanQian/VHAPv1.6.39Jun 18, 2026by Ed1s0nZ
AI Summary
The release unifies large tool output handling by removing legacy configuration parameters and implementing automatic truncation via Eino reduction for outputs exceeding 50KB. Saved outputs are now organized under `tmp/reduction/` based on project or conversation context.
Key Highlights
- Unified large tool-output handling logic
- Removed legacy configuration parameters
- Automatic truncation of outputs >50KB via Eino
- Organized file storage in `tmp/reduction/`
Breaking Changes
- Removed `large_result_threshold` configuration parameter
- Removed `result_storage_dir` configuration parameter
- Removed `query_execution_result` configuration parameter
New Features
- Eino reduction integration for large outputs
- Automatic file persistence for truncated results
Full Release Notes
本次更新统一了大工具输出处理：删除 large_result_threshold、result_storage_dir 和 query_execution_result；超过 reduction_max_length_for_trunc（50KB）的结果由 Eino reduction 生成 <persisted-output> 并落盘。落盘路径在 tmp/reduction/ 下按项目区分：已绑定项目为 projects/{项目ID}/，未绑定则为 conversations/{会话ID}/。
---
This update unifies large tool-output handling by removing large_result_threshold, result_storage_dir, and query_execution_result. Outputs above reduction_max_length_for_trunc (50KB) are truncated via Eino reduction into <persisted-output> and saved under tmp/reduction/, scoped by project (projects/{projectID}/ when a project is bound, otherwise conversations/{conversationID}/).