Sentry is an application monitoring platform for detecting, tracing, and fixing errors in production software. It collects errors, crashes, performance traces, logs, and session replays from application SDKs and groups them into issues with the stack traces, breadcrumbs, and request context needed to debug them. The platform runs as a hosted service at sentry.io or self-hosted from the same source.
For AI applications, Sentry adds Agent Tracing to follow LLM and tool calls, Seer for AI-assisted debugging with Agent Autofix and AI Code Review, and a Sentry MCP server that exposes issue data to coding agents. GitHub and Slack integrations and a Sentry CLI round out the workflow.
Features
- Error monitoring: grouped issues with stack traces, release tracking, and alerting
- Tracing and profiling: distributed traces across services, a trace explorer, and code-level profiling
- Logs and metrics: structured log ingestion and custom metrics alongside error data
- Session replay: recorded user sessions linked to the errors they triggered
- Agent tracing: spans for LLM calls, tool invocations, and agent steps
- Seer: AI debugging with Agent Autofix and AI Code Review
- Cron and uptime monitoring: scheduled-job and endpoint availability checks
- Size analysis: tracking of build and bundle sizes
- SDKs: JavaScript, Electron, React Native, Python, Ruby, PHP, Laravel, Go, Rust, Java and Kotlin, Objective-C and Swift, .NET, C and C++, Dart and Flutter, Perl, Clojure, Elixir, Unity, Unreal Engine, Godot, and PowerShell
- Integrations: GitHub, Slack, Sentry CLI, Sentry MCP, and OpenTelemetry
