Databricks Open Sources Omnigent to Put a “Meta-Harness” Above AI Agents. Databricks open-sources Omnigent, a meta-harness that composes, shares, and policy-controls multi-agent sessions across interfaces and sandboxes. Outcome engineers get a runnable orchestration layer for composing multi-agent workflows and enforcing policy boundaries — a concrete step toward Agentic Coordination and Gate controls (Principles 09, 15).

Ataraxy Labs’ Weave targets the merge conflicts AI agents create. Weave resolves Git conflicts by merging code as language-level entities and coordinates agents with a CRDT layer to avoid false conflicts. Use Weave to cut agent-induced merge churn and make artifact handoffs legible between human and agent contributors, improving Teamwork and Legible Landscapes (Principles 03, 06).

Depthfirst turns FFmpeg into a proof point for autonomous security agents. Depthfirst’s autonomous security agent finds 21 FFmpeg zero-days, proving agents that generate reproducible exploits change vulnerability economics. That outcome forces engineers to build immune systems and artifact-level verification into agent pipelines so produced outputs are auditable and safe (Principles 14, 16).

GLM-5.2 Is Out. GLM-5.2 launches as an open-source, 1M-context coding model with imminent API, chatbot services, and MIT licensing. Outcome engineers can leverage a long-context, open-weight coding model to run planning-plus-execution agents and shift important state out of fragile prompts into persistent artifacts (Principles 05, 06).

Cohere’s North Mini Code Turns Its Enterprise AI Pitch Toward Developers. Cohere releases North Mini Code, a 3B-parameter, 256K-context, Apache-2 coding model designed to run locally and power agentic developer workflows. Teams can run private, low-latency coding agents inside single-tenant infra or developer machines, making islanded agent deployments and reproducible delivery practical (Principles 07, 10).