graphsignal-sdk

Official

Instrument Python AI workloads with Graphsignal.

Authorgraphsignal
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Graphsignal SDK helps observe, profile, and debug AI inference workloads in Python by providing high-resolution profiling timelines, per-step tracing, and system metrics, enabling faster diagnosis and optimization.

Core Features & Use Cases

  • Auto-instrumentation for vLLM, SGLang, and PyTorch to capture traces, profiles, and metrics
  • Manual tracing and profiling APIs to annotate and measure custom code paths
  • GPU/CUPTI support for CUDA-based profiling and integration via graphsignal-run CLI

Quick Start

Install the Graphsignal SDK and configure it with your API key; then run your application with graphsignal-run to enable automatic instrumentation and observability.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: graphsignal-sdk
Download link: https://github.com/graphsignal/graphsignal-python/archive/main.zip#graphsignal-sdk

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.