perf: parallelize tool server init and reduce LLM retry overhead#139

Draft

wangbinluo wants to merge 4 commits intoMiroMindAI:mainfrom

wangbinluo:dev_wbl

Commits on Mar 17, 2026

perf: parallelize tool server init and reduce LLM retry overhead (#137)

wangbinluo
and
claude
committed
style: fix ruff formatting in manager.py

wangbinluo
and
claude
committed
revert: restore LLM retry parameters (max_retries=10, base_wait_time=30)

wangbinluo
and
claude
committed

Commits on Mar 18, 2026

perf: persistent MCP sessions — reuse subprocess across tool calls (#137)

wangbinluo
and
claude
committed