Explain A2CR MCP flows: WorkBaton serial handoff, WorkStash temporary memory, and WorkThreads multi-agent collaboration, with encryption and coordination details.
Create a shared memory space for multi-agent collaboration with customizable access permissions. Define allowed agents, tags, and description for fine-grained control.
Enables AI agents to share persistent memory via an MCP server using SQLite, supporting multi-tenant, categorized knowledge with TTL and semantic links, without requiring vector databases.
Sync AI agent memory files with a Git repository to maintain shared project context across multiple agents. Supports initializing, pushing changes, pulling updates, and checking sync status.
Audit multi-agent system communications to detect infinite delegation loops, privilege escalation, data leakage, and unauthorized handoffs across protocols like A2A, CrewAI, LangGraph, and AutoGen.
Retrieve and filter memory events such as shared-cache notifications from persistent storage. Use tags and timestamps to find only relevant, unconsumed updates.