numguard is an open-source verification layer for checking numerical claims in agent evaluations, leaderboards, and backtests. It produces signed reproducibility receipts and supports agent-callable access with prepaid credits and x402 pay-per-call metering.
numguard is an Agent evaluation & testing project. It focuses on preventing unverified numerical claims from entering agent evaluations, leaderboards, and backtests. numguard is an open-source project aimed at AI developers and researchers. The project is open source (MIT). It ships for the command line and API, and it can be self-hosted.
It is developed by ipezygj, and it first shipped in 2026. Development happens publicly on GitHub with 107 commits in the last 90 days. Among its 7 catalogued features are number verification, reproducibility receipts, and agent-callable checks. It exposes integrations via an MCP server and 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