2 boosters for "tracer" — open source, verified from GitHub, ready to install
Graphsignal observes inference workloads from a sidecar process — the profiler. It never shares a process with CUDA: the profiler watches the workload externally via CUPTI, OTLP/gRPC, Prometheus scraping, and NVML. Auto-instrumentation covers vLLM, SGLang, and PyTorch out of the box. Two install pat
Manage tasks and dependencies with Tracer CLI. Use for issue tracking, dependency management, finding ready work, and AI agent workflows.
