whereismyai
Apgar
Apgar is an open-source release health tool for AI agents that compares new deployments to a baseline from production traces.
Apgar monitors the first hours of every AI agent release to detect performance regressions compared to the previous release. It tracks tool errors, retries, loops, tool calls, answer length, cost, and latency. The tool analyzes traces you already collect on all traffic without requiring a dataset, stored prompt text, or completion text. Users send traces via OpenTelemetry or Langfuse, tag deployments with a release attribute, and receive a Slack message or webhook per release. Apgar runs on Docker Compose or Kubernetes using a Helm chart, and uses ClickHouse for data storage.
Apgar at a glance
- Best for: Software engineer
- Runs on: self-hosted, open source
- Pricing: Free
- Released: October 2026
- Category: AI coding tool
- Website: github.com/oimo-ai/apgar
Key features
- Compares new releases to a baseline from production traces
- Tracks tool errors, retries, and execution loops
- Monitors tool call volume, answer length, cost, and latency
- Sends alerts via Slack messages or webhooks
- Ingests traces over OpenTelemetry or from Langfuse
- Runs on Docker Compose and Kubernetes
Use it for
- Monitor AI agent release health and performance regressions
- Analyze production traces for tool errors and retry loops
- Track cost and latency changes across agent deployments
Alternatives to Apgar
- AG-UI — Standardizes agent-user interactions and generative UI for applications.
- Google ADK for Go — Apply standard software engineering practices to AI agent orchestration.
- jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
- APM — Standardize and reproduce AI agent configurations with dependency management.
- Aegis — It ensures AI coding agents work against verified baselines and proof.
- Git AI — It moves work to the server and handles concurrent tasks without standard Git bottlenecks.
- Antibody — It shares error fixes between parallel coding agents to save tokens and avoid redundant diagnoses.
- Genkit — Provides unified APIs and local developer tools for AI application workflows.
- OpenCode — Works with 75+ model providers, local models, and existing subscriptions
- abtop — Monitor multiple AI agent sessions and detect orphan ports from a single screen.
- CodeGraph — It eliminates file crawling, reducing AI agent tool calls by 88% and cutting costs.
- Goast — Pinpoints the root cause of production errors and generates pull requests instantly.
All Apgar alternatives