- STABILITY
- CONTROL
- VISIBILITY
- HUMAN
Short answer
How can I throttle message delivery with my AI?
Your AI uses QStash flow control keys to adjust parallelism and pause or resume message streams. This prevents your downstream services from being overwhelmed by sudden spikes in traffic.
Flow control outcomes
Manage your message pressure.
The AI acts as a buffer between your triggers and your infrastructure.
STABILITY
Protect downstream services
The AI adjusts flow control settings to ensure your databases or APIs don't crash under heavy load.
CONTROL
Pause and resume streams
Your agent can stop message delivery entirely or restart it once your services are ready to handle the volume.
VISIBILITY
Monitor active keys
The AI identifies which flow control keys are currently managing your message queues.
HUMAN
Manual intervention
You decide when to pin specific keys or change high-level logic that the AI shouldn't touch.
The workflow
What your AI does when <the thing arrives>.
The AI manages the pressure on your infrastructure through these moves.
Identify flow controls
The AI looks up existing flow control keys to see how your current message streams are being managed.
list_flow_control_keysInspect specific keys
It checks the settings of a single key to understand its current parallelism or limits.
get_flow_control_keyAdjust delivery speed
The AI pauses or resumes the flow to stop or start the message movement immediately.
pause_flow_controlresume_flow_controlLock settings
The AI can pin or unpin keys to ensure certain flow settings remain stable during operations.
pin_flow_controlunpin_flow_control
Try it
Copy these to start.
Use these prompts to begin managing your message flow.
Starting points
These are opening lines. Swap out the specific key names or service identifiers for your actual QStash setup.
Upstash QStash Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_qe2toOwnITMr2o4qUgpUtwgjBB41l4nVpwOjAbvw/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 — Upstash QStash capabilities are ready to use.
{
"mcpServers": {
"qstash-serverless-message-queue-mcp": {
"url": "https://edge.vinkius.com/vk_preview_qe2toOwnITMr2o4qUgpUtwgjBB41l4nVpwOjAbvw/mcp"
}
}
}List all my current QStash flow control keys.
Pause the flow control for the 'api-limit' key.
Check the settings for the 'worker-queue' key.
Resume the 'bulk-upload' flow control once the database is ready.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect QStash once, then ask.
Just link your account once. Your credentials stay encrypted, and you can start managing flows immediately through your agent.
Connect QStash to your AIFAQ
How this task behaves.
- 01
Can the AI delete my flow control keys?
No. The AI can only list, get, pause, resume, pin, or unpin keys.
- 02
Can the AI change the actual message content?
No. This MCP only manages the flow control settings, not the data inside the messages.
- 03
How does the AI know when to pause delivery?
The AI pauses delivery based on the specific instructions or conditions you provide in your chat.
- 04
Can the AI create new flow control keys?
No. The available tools are limited to managing existing keys.
- 05
Does the AI handle the actual message sending?
No. QStash handles the delivery; the AI simply manages the flow control settings that govern that delivery.
More questions about QStash? The Connector page answers them. See everything the QStash Connector can do
Connect Upstash QStash to Claude, Cursor, ChatGPT & more
