Verdict Inspect is a CLI tool that processes conversation exports (conversations.json) from ChatGPT, Claude, Cursor and similar LLMs. It produces structured reports analyzing key quality metrics including response drift, hedging language, refusal patterns, and overall conversation quality. Built for AI developers and evaluators who need to audit and improve the behavior of their LLM-powered applications.
Verdict Inspect is a LLM eval & observability project. Developers and evaluators lack an easy way to quantitatively assess LLM conversation quality, detect unwanted behaviors like hedging or refusals, and identify response drift from exported chat logs. It is built as an open-source project for developers. The project is open source (Apache-2.0). It ships for the web and the command line, and it can be self-hosted.
Cognifity builds and maintains Verdict Inspect, and it first shipped in 2026. The project is developed in the open on GitHub with 6 commits in the last 90 days. Among its 5 catalogued features are Chat Export Analysis, Drift Detection, and Hedging Detection. It exposes integrations via a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match