OBSERVABILITY

On August 4 Cloudflare launched the Agent Development Lifecycle: agent traces with session replay, @cloudflare/ci for CI/CD as Workflows, and local OpenTelemetry. A software factory is no longer just an idea, it is a platform product.

A new benchmark drops five frontier coding agents into a live OpenTelemetry microservice system with real Prometheus, Jaeger, and OpenSearch telemetry. Best RCA accuracy: 25.3% on Medium, 10.0% on Hard. Even Claude Fable 5 is far from oncall-ready.

A trending Codex SQLite WAL bug is a useful warning for every local coding agent: logs, disks, background processes, and telemetry paths need budgets too.

Open-source LLM engineering platform: tracing, evals, prompt management, and datasets. Self-hostable, OpenTelemetry-native, with 50+ framework integrations.

The latest Claude Code cache-burn debate is not just a quota complaint. It is a reminder that coding agents need cache-hit telemetry, spend ceilings, and repro-grade usage logs.

Error monitoring and performance tracing with release tracking, session replay, and first-class Next.js support.

Interactive TUI dashboard that shows exactly where your Claude Code and Cursor tokens are going, in real time.

AI agent work needs local observability. OpenTelemetry, OTLP, Vercel AI SDK telemetry, and lightweight trace viewers give developers receipts for model calls, tool use, latency, errors, and cost before anything goes to production.