- LIVE
- DEPLOY
- AUDIT
- HUMAN
Short answer
How can I update customer instance settings using my AI?
Your AI uses GraphQL to target specific customer instances and apply new configurations immediately. This allows you to push live updates to individual setups through a simple chat interface. You get direct control over deployment and instance state.
Configuration outcomes
Where changes land
What happens when you send a configuration command.
LIVE
Instant instance updates
The AI executes a query that changes settings for a specific customer. The change takes effect on that instance immediately.
DEPLOY
Direct configuration deployment
Your agent pushes new logic or parameters to a live environment. This bypasses manual configuration screens.
AUDIT
Configuration visibility
The AI reads current instance states via GraphQL. You can verify exactly what settings are active before making changes.
HUMAN
Manual intervention needed
If a query fails or a configuration is invalid, the AI stops. You must review the error and manually correct the data.
The workflow
What your AI does when the command arrives.
The AI moves from reading the current state to pushing the new configuration.
Fetch current state
The AI pulls the existing configuration data for a specific customer instance to ensure the context is correct.
run_graphqlDraft the update
The agent prepares the specific GraphQL mutation needed to change the target instance settings.
run_graphqlExecute deployment
The AI pushes the prepared configuration changes directly to the live customer environment.
deploy_instanceConfirm changes
The agent runs a final check to verify the new settings are active and correctly applied.
run_graphql
Try it
Copy these to start.
Use these prompts to trigger configuration changes.
Starting points
Swap the bracketed text for your actual customer IDs or specific setting names.
Prismatic Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_Jp8xQVyawqmk1kh1SNDCvDluadVzruFklDLm4lQl/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Prismatic capabilities are ready to use.
{
"mcpServers": {
"prismatic-mcp": {
"url": "https://edge.vinkius.com/vk_preview_Jp8xQVyawqmk1kh1SNDCvDluadVzruFklDLm4lQl/mcp"
}
}
}Show me the current configuration for customer instance [ID].
Update the [Setting Name] to [Value] for instance [ID].
Deploy the new configuration settings to [Customer Name].
Check if the [Feature Name] is enabled for instance [ID].
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Prismatic once, then ask.
Connect Prismatic to your AI client in one click. Your credentials stay encrypted, and you can start managing instances immediately.
Connect Prismatic to your AIFAQ
How this task behaves
- 01
Can the AI delete an entire customer instance?
No. The tools only allow for running queries and deploying configuration changes.
- 02
Does the AI need me to approve every change?
The AI performs the action as soon as you give the command, but you can always set your AI client to ask for confirmation before executing tools.
- 03
Can I use this to create brand new customers?
No. This connector is designed for managing and deploying configurations to existing instances.
- 04
What happens if a GraphQL query is malformed?
The tool will return an error to your AI, and the AI will report the failure to you instead of applying the change.
- 05
Can the AI see my Prismatic login password?
No. Your credentials are encrypted and managed by Vinkius; the AI only interacts with the tools provided.
More questions about Prismatic? The Connector page answers them. See everything the Prismatic Connector can do
Connect Prismatic to Claude, Cursor, ChatGPT & more
