| name | ops-verify-telemetry |
| description | Verify that OpenTelemetry traces are being collected and exported to X-Ray. Check trace health, find slow requests, investigate errors, and view service dependencies. |
| allowed-tools | ["Bash","Read"] |
Verify OpenTelemetry Telemetry
Verify that OpenTelemetry traces are being collected and exported to AWS X-Ray in the specified environment.
Target environment: $ARGUMENTS (expected: staging or production; default: staging)
Workflow
Step 1: Set Environment Variables
| Environment | Profile |
|---|
| staging | your-project-staging |
| production | your-project-production |
Step 2: Check AWS Session
aws sts get-caller-identity --profile <profile>
If expired: aws sso login --profile <profile>
Step 3: Check Trace Health
Run the health subcommand to get trace count and sample traces from the last 5 minutes:
bin/verify-telemetry health <profile>
This confirms that the ADOT sidecar is receiving spans from the application and forwarding them to X-Ray.
Step 4: Check Service Graph
Run the services subcommand to view the dependency map:
bin/verify-telemetry services <profile>
Verify that the your-project-<env> service appears in the graph and its downstream dependencies (MySQL, external APIs) are visible.
Step 5: Investigate Issues (Optional)
Based on user request or anomalies found in previous steps:
Slow requests (response time > 1s):
bin/verify-telemetry slow <profile>
5xx faults:
bin/verify-telemetry errors <profile>
Specific trace detail:
bin/verify-telemetry trace <trace-id> <profile>
Step 6: Report Results
Summarize findings in a table:
| Check | Result |
|---|
| AWS session active | Yes/No |
| Traces collected (last 5 min) | Count |
| Service graph populated | Yes/No (list services) |
| Slow requests (last hour) | Count |
| 5xx faults (last hour) | Count |
If any check fails, provide the specific error output and a recommended next step.
Execution
Verify telemetry now for the specified environment.
