Short version: .cursorrules loads Rediacc context into Cursor’s AI; the terminal lets it run rdc commands against your actual machines.
Quick Setup
- Install the CLI:
curl -fsSL https://www.rediacc.com/install.sh | bash - Copy the AGENTS.md template to your project root as
.cursorrules - Open the project in Cursor
Cursor reads .cursorrules at startup. Here’s the thing: context window limits apply, so keep the file focused on your actual machines and repos rather than generic boilerplate.
.cursorrules Configuration
Create .cursorrules in your project root with the Rediacc infrastructure context. See the full AGENTS.md template for a complete version.
The key sections to include:
- CLI tool name (
rdc) and installation - Common commands with
--output jsonflag - Architecture overview (repository isolation, Docker daemons)
- Terminology rules (adapters, not modes)
Terminal Integration
Cursor can execute rdc commands through its integrated terminal. Common patterns:
Checking Status
Ask Cursor: “Check the status of my production server”
Cursor runs in terminal:
rdc machine status prod-1 -o json
Deploying Changes
Ask Cursor: “Deploy the updated nextcloud config”
Cursor runs in terminal:
rdc repo up nextcloud@prod-1 --yes
Viewing Logs
Ask Cursor: “Show me the recent mail container logs”
Cursor runs in terminal:
rdc repo logs mail@prod-1 -c mail-postfix --lines 100
Workspace Settings
For team projects, add Rediacc-specific Cursor settings to .cursor/settings.json:
{
"terminal.defaultProfile": "bash",
"ai.customInstructions": "Use rdc CLI for all infrastructure operations. Always use --output json when parsing results."
}
Tips
- Cursor’s Composer mode works well for multi-step infrastructure tasks
- Use
@terminalin Cursor chat to reference recent terminal output - The
rdc --help-allcommand gives Cursor a complete command reference - Combine
.cursorruleswith aCLAUDE.mdfile for maximum compatibility across AI tools