Ask questions about your production systems in natural language. NOFire is trained to understand service names, time ranges, and technical concepts without requiring specific syntax.
Works in:
- NOFire MCP Server and your IDE integration
- NOFire Dashboard (chat interface)
- NOFire Slack App (@mention the bot)
Time Windows
NOFire always operates in a time window. If you don’t specify one, it defaults to the last 1 hour.
How to specify time:
- Relative: “in the last 30 minutes”, “last 2 hours”, “since yesterday”
- Absolute: “between 2pm and 3pm today”, “from 9:00 AM to 11:00 AM”
- Specific moment: “around 3:24 pm ET on April 10”
Time-scoping your prompts ensures NOFire analyzes the correct telemetry data.
Prevent Issues
Catch problems before they become incidents.
Deployment Safety
can I safely deploy checkout-api today?
what's the risk score for deploying payment-service now?
are there any incidents in the last 24 hours that would block a deployment?
Alert Coverage
are alerts properly configured to catch payment-service failures?
will checkout-api alert me if it degrades?
Instrumentation Gaps
is frontend-service missing critical instrumentation?
are we collecting the right signals from recommendation-service?
what metrics are available for checkout-api?
Configuration Validation
are health checks configured correctly for frontend-service?
are resource limits set appropriately for payment-service?
Quick Insights
Fast access to metrics, logs, and traces for immediate understanding.
give me the cpu and memory trend for recommendationservice
show me the error rate for checkout-api
what is the latency for payment-service in the last 2 hours?
Logs & Errors
show error logs from frontend-service in the last 30 minutes
give me the logs for api-d5d94d6
were there any 5xx errors around 5pm yesterday?
Service Health
is my cluster otel-demo healthy?
is recommendationservice running?
show me the slowest trace in checkout-api
Root Cause Analysis
Understand why things broke and what triggered issues.
Service Failures
find the root cause of errors in payment-service
why did checkout-api start failing 10 minutes ago?
what caused the latency spike in recommendation-service?
Change Impact
show me what changed before frontend-service degraded
what changed on checkout-api in prod before the latency spiked at 2:15 pm?
is checkout-api impacting downstream dependencies?
Configuration Issues
why isn't prometheus scraping data from my services?
why are my pods constantly restarting?
explain trace e97f261d4fccfdbfb19be22e5174ab5e that happened 2 hours ago
Optimize Costs
Reduce observability spend without losing visibility.
Cost Analysis
how much is payment-service costing in observability?
what's our total observability spend?
Reduce Waste
what metrics can we safely remove from checkout-api?
show me high-cardinality metrics in recommendation-service
is checkout-api generating unnecessary log data?
Retention & Sampling
can we reduce log retention costs?
can we reduce trace costs for frontend-service?
Spans & Traces
For performance bottlenecks and distributed system debugging.
check the spans for error patterns in checkout-api
explain why traces in checkout-api have been slower than 2 minutes
show me traces with high latency in the last 10 minutes
Events & Changes
For change detection and deployments.
were there any change events on checkout-api in the last day?
check kubernetes events on the production namespace
Tips for Better Prompts
-
Be specific about time ranges
- “in the last 30 minutes”
- “between 2pm and 3pm today”
-
Name services naturally
- “checkout-api” or “checkout api”
- “recommendation service”
-
Ask follow-ups in the same thread
- NOFire maintains context from previous messages
-
Include symptoms when investigating
- “why are error rates spiking in payment-service?”
- “what’s causing slow responses in the API?”
-
Use specific trace/error IDs when available
- “explain trace e97f261d4fccfdbfb19be22e5174ab5e”
- “show me logs for pod frontend-xyz123”
