[Question]: Concurrency issues with Feishu bot + Manager-Agents setup (serial execution & context cross-contamination)

## Question or topic

I am testing the concurrency capability of the CoPaw framework when deployed as a Feishu bot. My setup uses a Manager Agent to coordinate two Worker Agents for task execution. However, during multi-user concurrent testing, I encountered issues:

Tasks are executed serially (one after another) instead of in parallel.
Results from different users are mixed up, with severe context leakage and cross-contamination between sessions.
I need to confirm whether this is a current limitation of CoPaw or an error in my configuration.

## Context

CoPaw Version: [v1.0.0.post3]
Deployment Channel: Feishu Bot
Agent Structure: 1 Manager Agent + 2 Worker Agents (task-execution roles)
Workflow: User sends a task via Feishu → Manager Agent dispatches to appropriate Worker → Worker returns result → Manager aggregates and replies
Testing Method: Two different Feishu users send tasks simultaneously to the bot