Forkline is a replay-first tracing and diffing library for agentic AI workflows that lets you deterministically reproduce, fork, and compare agent runs to find exactly where behavior diverged.
-
Updated
Feb 27, 2026 - Python
Forkline is a replay-first tracing and diffing library for agentic AI workflows that lets you deterministically reproduce, fork, and compare agent runs to find exactly where behavior diverged.
Prompt and response tracing for LLM workflows
OKI TRACE: Local LLM observability. See step-by-step, layer-by-layer what your AI thinks. Logit Lens & Attention for HuggingFace models.
Stop guessing why your MLX model outputs garbage. Triage in 30 seconds — no model load required.
Add a description, image, and links to the llm-debugging topic page so that developers can more easily learn about it.
To associate your repository with the llm-debugging topic, visit your repo's landing page and select "manage topics."