Contextual Policies in Omnigent: Using Session State to Better Govern AI Agents
Databricks, Tuesday, July 7th, 2026
Databricks introduces Omnigent, an open-source meta-harness enabling contextual, session-aware policies for AI agents.
Databricks introduced Omnigent, an open-source meta-harness that enables contextual policies for AI agents.
These policies track session history and state to make dynamic decisions about agent actions rather than applying static rules.
The framework supports popular coding agents like Claude Code and Codex, offering capabilities such as keeping an agent's access scoped to the right documents, monitoring cumulative risk scores, and enforcing budget constraints based on what the agent has already done in a session.