Hew ships with hew-observe, a terminal dashboard for inspecting running actor systems. Enable profiling with the HEW_PPROF environment variable, then attach the observer:

HEW_PPROF=:6060 hew run myapp.hew
hew-observe              # in another terminal

Overview tab — system-wide metrics at a glance:

hew-observe overview tab showing system metrics, message rates, active workers, and memory usage

The overview shows actor count, message throughput (msg/s), memory usage, and live charts for messages, workers, and allocations.

Actors tab — per-actor detail:

hew-observe actors tab showing per-actor state, message count, mailbox depth, and CPU time

Each actor shows its state (running, idle, blocked, crashed), message count, average dispatch latency, current mailbox depth, high-water mark, and cumulative CPU time. Sort by any column with s, filter with /.

The observer also has tabs for Supervisors (tree hierarchy), Crashes (fault log), Cluster (distributed topology), Messages (sequence diagrams), and Timeline (event history).

Try hew-observe --demo to explore the dashboard with synthetic data.