AgentCrew and mode-manager-mcp

These are ecosystem siblings—mode-manager-mcp provides a VS Code-integrated instruction and memory management layer that could enhance the multi-agent orchestration capabilities of a chat application like AgentCrew by standardizing agent behavior and context across MCP servers.

AgentCrew
53
Established
mode-manager-mcp
50
Established
Maintenance 13/25
Adoption 10/25
Maturity 9/25
Community 21/25
Maintenance 6/25
Adoption 10/25
Maturity 18/25
Community 16/25
Stars: 161
Forks: 33
Downloads:
Commits (30d): 0
Language: Python
License: Apache-2.0
Stars: 10
Forks: 6
Downloads: 215
Commits (30d): 0
Language: Python
License: MIT
No Package No Dependents
No risk flags

About AgentCrew

saigontechnology/AgentCrew

Chat application with multi-agents system supports multi-models and MCP

Supports both **transfer mode** (sequential task handoff with full context carry-over) and **delegate mode** (parallel task dispatch returning results as tool output), alongside MCP protocol integration, web search, file editing, browser automation, and memory systems. Built on an agent specialization architecture where tasks route to specialized agents, with parallel execution of safe tools via `asyncio.gather`. Integrates Claude, GPT, Gemini, GitHub Copilot, and custom providers, plus exposes agents as HTTP services via the A2A protocol for distributed multi-system collaboration.

About mode-manager-mcp

NiclasOlofsson/mode-manager-mcp

MCP Memory Agent Server - A VS Code chatmode and instruction manager with library integration

Implements a three-tier memory architecture (personal, workspace, and language-specific) using persistent Markdown files with YAML frontmatter, enabling context-aware AI assistance across VS Code projects. Operates as an MCP server via stdio transport, integrating directly with GitHub Copilot through natural language commands—users store knowledge by simply saying "remember" without configuration overhead. Automatically optimizes and consolidates memories to maintain efficiency while sharing team knowledge through `.github/instructions` directories for distributed onboarding and consistency.

Scores updated daily from GitHub, PyPI, and npm data. How scores work