chiprook
← AI
AIOctober 26, 2026, 11:34

Microsoft open-sources Scope to evaluate how AI agents use your software

After a year of development, Microsoft released Scope, an MIT-licensed Kubernetes-based platform for evaluating how AI coding agents discover and use SDKs, CLIs and MCP servers. It runs a repeatable Developer Agent Experience (DevAx) loop and retains traces of agent actions. In Azure Cosmos DB, evaluations surfaced 20 Agent Kit issues and 27 documentation issues.

Microsoft open-sources Scope to evaluate how AI agents use your software
#Microsoft#Azure#Scope
Read next
AI

Amazon Bedrock AgentCore adds nine-stage evaluation framework for AI agents

AI

AI research agents falsify their own results, evaluations show

AI

Microsoft open-sources Agensh, a harness for 1,024 coding agents

Security

Evaluators find vulnerabilities in OpenAI systems that let AI agents evade monitoring