From df72fa75ca3aed85ea2d7ae4da0170efb2c6da6d Mon Sep 17 00:00:00 2001 From: Alex Akimov Date: Tue, 13 Oct 2026 22:31:56 +0200 Subject: [PATCH] Update README.md --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 30dc155..72cd6cd 100644 --- a/README.md +++ b/README.md @@ -83,6 +83,7 @@ Collection of awesome Python resources for testing and generating test data. - [agent-convergence-scorer](https://github.com/hermes-labs-ai/agent-convergence-scorer) - A CLI and Python library that scores how lexically similar N agent or LLM outputs are: exact-match rate, Jaccard token overlap, divergence point, and a composite convergence score over any list of runs. An eval primitive for measuring reproducibility and fan-out collapse. Zero deps. - [mcp-server-fuzzer](https://github.com/Agent-Hellboy/mcp-server-fuzzer) - A comprehensive fuzzing tool designed specifically for testing Model Context Protocol (MCP) servers. It supports both tool argument fuzzing and protocol type fuzzing across multiple transport protocols. +- [mcpscore](https://github.com/mcp-box/mcpscore) - Audits a running MCP server and scores it 0-100 for protocol conformance, tool and schema quality, and security posture, with a suggested fix for every failed check. Deterministic, runs from the CLI or as a GitHub Action with SARIF output. MIT. - [OrcaReplay](https://github.com/Continuum-AI-Corp/OrcaReplay) - Records a Python agent's calls to its model provider from outside the process, then serves the recording back so the same run happens again with the provider unreachable and no key. Turns a failed session into a regression test. Apache-2.0. - [Sabot](https://github.com/Jott2121/sabot) - Plants controlled faults inside running multi-agent pipelines (LangGraph, CrewAI, AutoGen) and scores whether each pipeline's own reviewer and guardrail stages detect them. Pre-registered spec, deterministic adjudication, published raw traces. - [spantruth](https://github.com/Plexito-de/spantruth) - Test suite for Python LLM telemetry libraries such as OpenLLMetry and OpenInference. A local fake provider serves OpenAI and Anthropic responses, including streams, prompt-cache hits, 429s and dropped connections, and spantruth checks the token counts, models and errors on the resulting spans, and whether each span ends, against what was sent. Apache-2.0.