agentharness-verifier is an open-source tool for independently verifying and auditing claims made by coding agents. It provides automated testing and audit trails to ensure the reliability and trustworthiness of agent outputs.
In the LLM eval & observability space, agentharness-verifier takes a focused approach. It focuses on ensuring coding-agent outputs are independently verified and auditable for correctness and trust. agentharness-verifier is an open-source project aimed at AI developers and teams building or evaluating coding agents. agentharness-verifier is open source under the Apache-2.0 license. It runs on the command line, and it can be self-hosted.
Behind agentharness-verifier is Fabio Scialanga, and it first shipped in 2026. Development happens publicly on GitHub with 72 commits in the last 90 days. Among its 5 catalogued features are claim verification, audit trails, and coding agent support.
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