Traceway

Use it when

• You want logs, traces, metrics, exceptions, and session replay in one self-hosted MIT-licensed system instead of gluing Prometheus, Loki, Tempo, and a replay tool together
• Your coding agent should go from a bug report to root cause by querying production exceptions, logs, and endpoint latency through a read-only CLI

What it solves

Not the fit when

• Profiling is experimental. AI observability integrations currently list OpenRouter and OTel-compatible gateways. Standalone server mode needs ClickHouse and PostgreSQL; embedded SQLite/DuckDB modes trade dashboard headroom for zero dependencies.
• synthetic uptime monitoring
• load and stress testing
• on-call paging rotation management

Install

git clone https://github.com/tracewayapp/traceway && cd traceway && docker compose up -d

Invoke

Open the dashboard at http://localhost and point any OTel SDK at http://localhost/api/otel/v1/traces (or /metrics, /logs); npx skills add tracewayapp/traceway installs agent skills and a read-only query CLI

Alternatives

No reviewed alternatives recorded yet.