Skip to main content

Examples

Explore how to use the WhaTap MCP server through natural language query examples and real conversation flows.

Frequently used queries​

SituationExample query
Getting started"List all my WhaTap projects"
Check server status"Show CPU usage for project 12345 over the last 5 minutes"
Check APM anomalies"Detect APM anomalies with high sensitivity"
Incident response"Summarize error status for the last 5 minutes"
Data exploration"What data can I query in this project?"
K8s check"Show pod status for project 33194"
Service relationships"Show service topology"
OpenMetrics exploration"Show available OpenMetrics metrics for project 33194"
PromQL creation"Create a PromQL query for CPU usage by Pod"
Run saved query"Run the saved 'CPU by Pod' query"
Agent installation"Install the server monitoring agent for project 12345"

Natural language query examples​

Sample queries for common monitoring tasks. Choose or adapt the queries that match your needs.

Project exploration​

List all my WhaTap projects
Show me the details of project 12345
What agents are in project 12345?

Server monitoring​

Show me server CPU usage for the last hour
Are there any servers with memory usage over 80%?
Check disk usage
Show me the top 5 servers by CPU usage

APM (Application Performance)​

Show me TPS trends for the last 5 minutes
What is the average response time?
Have any transaction errors occurred?
Check the status of currently active transactions

Kubernetes​

Check Pod status
Show CPU and memory usage by node
Are there any Warning events in recent cluster events?

Database​

Show me the DB instance list
Check current active sessions

Anomaly detection​

Find agents with anomalies in the last 5 minutes
Run anomaly detection with high sensitivity for the past hour

Service topology​

Show service connections
Are there any network bottlenecks?

OpenMetrics / PromQL​

Show available OpenMetrics metrics for project 33194
Create a PromQL query for CPU usage by Pod
Run the saved 'CPU by Pod' query

Agent installation​

How do I install a monitoring agent for project 12345?
Install the Java APM agent
Show me how to set up the Kubernetes agent
Install PostgreSQL DB monitoring
Install the infrastructure agent on a Debian server

Composite queries​

Do a morning health check.
Check server status across all projects,
and analyze whether there are any anomalies in APM projects.

Conversation examples​

The conversations below are examples. Actual responses will vary depending on your project's monitoring data.

Example 1. Morning server check​

User: Do a morning check. Summarize server CPU and memory status.

AI: I checked 3 servers.
    - web-01: CPU 45%, Memory 62% (Normal)
    - web-02: CPU 87%, Memory 71% (CPU warning — exceeded recommended 70%)
    - db-01:  CPU 23%, Memory 55% (Normal)

    web-02's CPU is high at 87%. Would you like to check its processes?

Example 2. APM anomaly detection​

User: Check if there are any anomalies in the Java APM project.

AI: I analyzed 5 agents.

    Anomaly detected (1 instance):
    - api-server-03: Response time spike (avg 120ms → max 2.3s, +4.2σ)

    The remaining 4 agents are within normal range.
    Would you like to check the transaction details for api-server-03?

Example 3. Incident response​

User: I just got an error alert. Summarize the error situation for the last 5 minutes.

AI: Here is the transaction error summary for the last 5 minutes.

    | Agent  | TPS  | Errors | Error rate |
    | ---    | ---  | ---    | ---        |
    | api-01 | 45.2 | 23     | 10.2%      |
    | api-02 | 42.8 | 2      | 0.9%       |
    | api-03 | 44.1 | 0      | 0%         |

    api-01 has an error rate of 10.2%, which is abnormal.
    Would you like to check the response time trend and Active TX as well?

Example 4. Executive report data​

User: Summarize this week's response time and error rate trends across all services.

AI: Here is the service status for the last 7 days.

    - Average response time: 142ms (-8% vs. previous week)
    - Average error rate: 0.3% (no change vs. previous week)
    - Average TPS: 1,240/s (+12% vs. previous week)

    Overall the service is in good shape. Despite the TPS increase,
    response time has actually decreased, indicating improved performance.
Last updated on