Can I customize the latency alert thresholds?

Yes, the provided Prometheus AlertManager rules are templates that allow you to easily adjust the P95 latency and error rate percentages to fit your specific SLAs.

How does this skill handle error tracking?

It implements structured error counters and instrumented clients that categorize Retell AI errors by type, which can then trigger automated alerts in AlertManager.

Does this require specific libraries?

Yes, it leverages popular industry standards including 'prom-client' for metrics, '@opentelemetry/api' for tracing, and 'pino' for structured logging.

What monitoring backends are compatible with this skill?

This skill is optimized for Prometheus for metrics, OpenTelemetry for tracing, and Grafana for visualization, though the patterns can be adapted for other compatible tools.

Retell AI Observability

Name: Retell AI Observability
Author: Brmbobo

byBrmbobo

0•

분석 및 모니터링

Implements comprehensive monitoring, distributed tracing, and automated alerting for Retell AI voice integrations.

The Retell AI Observability skill provides a standardized framework for monitoring the health and performance of Retell AI voice applications. It automates the implementation of Prometheus metrics for request tracking, OpenTelemetry for distributed tracing, and structured JSON logging for deep production debugging. By offering pre-configured AlertManager rules and Grafana dashboard templates, it enables developers to proactively detect latency spikes, error rate increases, and integration failures, ensuring high reliability for voice-driven AI services.

주요 기능

01OpenTelemetry integration for end-to-end distributed tracing across services

02Grafana dashboard queries for real-time visualization of integration health

030 GitHub stars

04Pre-defined AlertManager rules for error rates and P95 latency thresholds

05Structured JSON logging strategy using Pino for consistent log analysis

06Automated Prometheus metrics for API request counts and latency histograms

사용 사례

01Debugging intermittent latency issues in real-time voice interactions

02Configuring automated alerts for Retell AI service degradation or downtime

03Setting up production-grade monitoring for voice-based AI agents

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add brmbobo/web2podcast retellai-observability

For use in Claude.ai and ChatGPT

Download Skill