Turn specs into evals for any agent with ASSERT
ID: f9d8f33d-1cdb-5787-b178-f6ab3e7d382b
STIX ID: report--f9d8f33d-1cdb-5787-b178-f6ab3e7d382b
Feed Name: Microsoft Security
Date Published: 2026-06-10
Date Updated: 2026-06-11
Author: Mehrnoosh Sameki, Sandeep Atluri, Minsoo Thigpen and Abby Palia
...
...
ASSERT is an open-source, MIT-licensed framework for turning natural-language behavior specifications into executable evaluation pipelines: it systematizes behavior specs into taxonomies, generates stratified test cases, runs them against models or agents (recording full traces and tool use), and scores results with rationales and policy citations to aid inspection and iteration.
Your team is not currently subscribed to this feed. You must subscribe to it in order to see this post.