Alloy MCP server
{
"mcpServers": {
"alloy": {
"type": "http",
"url": "https://api.alloy.cx/mcp"
}
}
}{
"mcpServers": {
"alloy": {
"type": "http",
"url": "https://api.alloy.cx/mcp"
}
}
}{
"mcpServers": {
"alloy": {
"type": "http",
"url": "https://api.alloy.cx/mcp"
}
}
}{
"servers": {
"alloy": {
"type": "http",
"url": "https://api.alloy.cx/mcp"
}
}
}{
"mcpServers": {
"alloy": {
"type": "http",
"url": "https://api.alloy.cx/mcp"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 1.0.0
What it does
Alloy is an MCP server that connects AI agents to a shared workspace where teams and AI agents collaborate on artifacts, knowledge, and access to external systems. It enables agents to read organizational context, write durable artifacts, and coordinate work across multiple tools and other agents.
- Read shared organizational knowledge, decisions, and communication history before responding to questions.
- Write proposals, research, summaries, and reusable notes back to Alloy so other agents and humans can access them.
- Access approved business systems like CRMs, support tools, and databases through Alloy-controlled MCP without configuring each integration separately.
Tools this server exposes
| Tool | Does |
|---|---|
| work-with-alloy | Core skill that teaches agents to read shared knowledge from Alloy, write durable artifacts, create workflows, and spawn autonomous agents. |
| review-alloy-operations | Skill for auditing teammates, instruction skills, workflows, MCPs, and recent activity within Alloy. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.