Emisar MCP server
{
"mcpServers": {
"emisar": {
"type": "http",
"url": "https://emisar.dev/api/mcp/rpc"
}
}
}{
"mcpServers": {
"emisar": {
"type": "http",
"url": "https://emisar.dev/api/mcp/rpc"
}
}
}{
"mcpServers": {
"emisar": {
"type": "http",
"url": "https://emisar.dev/api/mcp/rpc"
}
}
}{
"servers": {
"emisar": {
"type": "http",
"url": "https://emisar.dev/api/mcp/rpc"
}
}
}{
"mcpServers": {
"emisar": {
"type": "http",
"url": "https://emisar.dev/api/mcp/rpc"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 0.49.0
What it does
emisar exposes declared infrastructure actions to MCP-capable agents through a control plane and outbound-only runner on Linux hosts. It lets you set policy for which actions require approval, which are denied, and which run freely, with full audit trails of every execution attempt.
- Dispatch host operations like uptime checks or package updates through an AI agent while requiring approval for risky changes.
- Audit every action an agent attempts, including denied and pending requests, with hash-chained journals on the host and control-plane records.
- Add new capabilities to agents by installing action packs from a catalog without reconfiguring MCP servers or redeploying clients.
Tools this server exposes
| Tool | Does |
|---|---|
| linux.uptime | Check host uptime. |
| shell | Break-glass staging-only pack for arbitrary command execution; critical-risk, default-denied, never suggested for production. |
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.