Configures and manages isolated or shared memory stores for multi-agent and team-based AI workflows.
The Multi-Agent Memory Manager skill provides a comprehensive interactive wizard to set up and fine-tune how Claude Code interacts with memory across different environments. It allows users to define unique agent identifiers, choose between unified or separate storage strategies, and configure team-wide memory sharing permissions. By detecting existing system states and providing automated validation for storage paths and identifiers, it ensures that memory stores remain organized, searchable, and secure, whether you are a solo developer using multiple AI tools or part of a collaborative engineering organization.
Características Principales
01Interactive wizard for single-user, multi-agent, or team mode configuration.
02Granular cross-agent query permissions and team sharing controls.
034 GitHub stars
04Automated state detection of existing memory-daemon configs and agent IDs.
05Flexible storage strategies including unified tagged stores or isolated environments.
06Built-in validation for agent ID formats and storage path writability.
Casos de Uso
01Standardizing agent identifiers across multiple machines to maintain a consistent interaction history.
02Isolating memory between different AI agents like Claude Code and Cursor to prevent context pollution.
03Configuring a shared team memory store to allow developers to query previous solutions across the organization.
