agentgap is an open-source, read-only auditing tool for evaluating the reliability and safety of AI agent repositories. It uses static analysis and sandbox-oriented checks to help developers identify risks without changing repository contents.
agentgap is a LLM eval & observability project. It focuses on auditing AI agent repositories for reliability and safety issues without modifying their code. It is built as an open-source project for AI agent developers and repository maintainers. agentgap is open source under the MIT license. agentgap is available on the command line, and it can be self-hosted.
It is developed by Ayush Kumar, and it first shipped in 2026. Development happens publicly on GitHub with 32 commits in the last 90 days. Among its 6 catalogued features are read-only auditing, static analysis, and safety checks.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do