GitHub shipped a substantial update to Copilot for JetBrains IDEs on August 11 that closes much of the feature gap with the VS Code extension. The two most consequential additions are Copilot Memory, which lets the agent retain and recall useful context across separate chat sessions instead of forcing developers to re-explain project details and preferences every time they open a new conversation, and native support for Ollama as a bring-your-own-key provider, meaning teams can now point Copilot's JetBrains integration at models running entirely on local infrastructure rather than a cloud API. That second change matters more than it might look: for regulated industries, air-gapped environments, or teams simply trying to control inference costs, the ability to swap in a self-hosted model without leaving the IDE removes a real adoption blocker that previously pushed some organizations toward alternative tooling. The release also expands Codex workflow support — sessions are now visible in agent debug logs, and Codex-based workflows pick up updated permission modes and support customization through instructions and skills files, bringing them closer to parity with how Copilot's own agents are configured. On the governance side, administrators get new server-based enterprise managed settings covering plugin availability, MCP server access, permission bypass behavior, and OpenTelemetry export, extending the kind of centralized AI-tooling governance that data platforms like Databricks and Snowflake have been building into their own AI gateways this same month, but now at the IDE and developer-tool layer instead of the data layer. Smaller quality-of-life fixes include automatic Copilot CLI installation from JetBrains' integrated terminal on macOS, Linux, and Windows, plus reliability fixes across MCP execution, terminal output rendering, and diff-based editing. For engineering leads evaluating AI coding assistants across a mixed IDE fleet, this release is a meaningful argument for consolidating on Copilot rather than running separate tools for VS Code versus JetBrains users.