GitHub Copilot for JetBrains Adds Cross-Session Memory and Ollama Model Access
The August 11 update combines persistent context, a local-model provider, Codex logs, and managed controls, so users should verify each data boundary separately.
Key takeaways
GitHub updated Copilot for JetBrains on August 11, 2026 with Copilot Memory across agent chat sessions and Ollama as a bring-your-own-key model provider. The release also expands Codex debug logging, permission modes, instruction and skill customization, and Copilot CLI setup from integrated terminals. Users can avoid repeating project context and select local models from JetBrains, but the feature names do not prove that every request stays offline or that every remembered item is appropriate. Before using sensitive repositories, inspect the Memory toggle and organization policy, test recall and deletion with harmless preferences, verify the Ollama endpoint and model, and review network behavior, logs, MCP approvals, terminal permissions, and a documented path back to default settings.
# GitHub Copilot for JetBrains Adds Cross-Session Memory and Ollama Model Access
August 13, 2026
Direct answer
Validate Memory and Ollama as separate controls. Test what Memory stores and recalls, then verify the Ollama endpoint and network path. Do not expose production secrets until both behaviors are evidenced.
Fact sources
GitHub announced cross-session Copilot Memory for JetBrains on August 11, 2026.
The same update adds Ollama as a BYOK provider with provider configuration and model selection in JetBrains.
Codex sessions now appear in agent debug logs, with updated permission, instruction, skill, and CLI workflows.
Five steps before enabling Memory and Ollama
- Confirm the Memory toggle, organization policy, and current JetBrains plugin version.
- Test save, recall, and disable behavior with project preferences that contain no sensitive data.
- Configure an Ollama endpoint and test model, then verify the actual request path and failure mode.
- Review Codex debug logs, permission prompts, MCP approvals, and terminal auto-approval rules.
- Document how to clear memory, revoke model configuration, remove logs, and restore the default provider.
Why it matters
Persistent context and local-model access both change where data lives and how failures appear. Convenience labels cannot replace an observed data-flow and policy check.
Impact for ordinary AI users
JetBrains users gain continuity and another model choice, while taking on hardware, model-quality, logging, and policy checks. Organization users may also face managed restrictions on plugins and permissions.
Related tools and tutorials
Start with one reversible task, verify version, permissions, cost, and logs, then record the result in the team runbook.
AI software and tools · AI account and cost services · AI skill tutorials · AI frontier news
FAQ
Does Copilot Memory retain every chat?
GitHub describes retaining useful information; verify the product toggle and organization policy instead of assuming identical behavior everywhere.
Does Ollama make the complete workflow offline?
Not necessarily. Verify the endpoint, plugin network traffic, model downloads, and any other Copilot services involved.
Why inspect Codex debug logs?
They help locate session and permission behavior, but the logs themselves must also be handled as potentially sensitive.
Source links
- GitHub Changelog: Copilot memory and Ollama in JetBrains (2026-08-11)
- GitHub Docs: About GitHub Copilot Memory
- JetBrains Marketplace: GitHub Copilot versions
- Ollama official download
What this means for everyday users
Record Memory policy, test data, endpoint, model version, network path, logs, permission mode, and cleanup steps.
Related reading
GitHub adds enterprise controls for Copilot agent commands, files, and network access
GitHub released enterprise-managed permissions for Copilot agent operations on September 9. Administrators can centrally set shell commands, file reads and writes, and access to network domains to blocked, approval required, or allowed without a prompt. User preferences, workspace settings, automatic approval, and earlier approvals cannot make the enterprise policy less restrictive. GitHub says the controls are generally available in the Copilot app, Copilot CLI, and Visual Studio Code sessions that use Agent Host for Copilot Business and Enterprise customers. Security and platform teams should begin with a minimum-permission baseline, test representative repositories, and expand only the operations that have a clear owner, audit trail, and rollback path.
AWS connects AgentCore evaluations to GitHub Actions for pre-merge agent regression gates
AWS published a reference workflow on September 8 for integrating Amazon Bedrock AgentCore Evaluations with GitHub Actions. The pipeline deploys a development agent and an OAuth-protected MCP server, invokes representative prompts, collects OpenTelemetry traces, and scores behavior before allowing a pull request to proceed. AWS describes built-in dimensions such as helpfulness, correctness, goal success, tool selection, tool parameters, and trajectory order, with custom and code-based evaluators available. The post compares three authentication patterns: evaluating stored traces, using a pre-authorized test user, or issuing machine-to-machine credentials for CI. Its implementation uses the third pattern. Teams should validate that CI credentials cannot escape their intended environment, that evaluation samples cover important failures, and that score variance, latency, evaluator cost, and rollback behavior are visible before treating the threshold as a release gate.
GitHub Copilot Customize Tab Is Generally Available for Team Agent Workflows
GitHub Copilot Customize Tab Is Generally Available for Team Agent Workflows. The official source dated August 2026 describes a concrete product, research, or governance change rather than a universal guarantee. This article separates what is available now from preview or planned access, then translates the change into one ordinary-user task: configuring team agent behavior in Copilot and validating results with a small task. Before using it, readers should verify account eligibility, workspace permissions, data boundaries, model or service cost, human review, audit logs, and rollback. A small reversible pilot with explicit acceptance checks is safer than copying a headline result or assuming that a new integration can publish, merge, or make decisions without approval. The source set is linked so teams can recheck availability and scope when the product changes.
GitHub Makes Global Model Policy Generally Available for Copilot
GitHub Makes Global Model Policy Generally Available for Copilot. The official source dated August 2026 describes a concrete product, research, or governance change rather than a universal guarantee. This article separates what is available now from preview or planned access, then translates the change into one ordinary-user task: standardizing Copilot model access rules across a team while preserving evidence of policy changes. Before using it, readers should verify account eligibility, workspace permissions, data boundaries, model or service cost, human review, audit logs, and rollback. A small reversible pilot with explicit acceptance checks is safer than copying a headline result or assuming that a new integration can publish, merge, or make decisions without approval. The source set is linked so teams can recheck availability and scope when the product changes.
GitHub Code Quality Adds Audit-Log Tracking for Enablement Changes
Review the official scope, availability, ordinary-user task, permissions, cost, review, and rollback checks for GitHub Code Quality Adds Audit-Log Tracking for Enablement Changes.
GitHub Copilot Enters Slack Public Preview for Shared Coding Work
Review the official scope, availability, ordinary-user task, permissions, cost, review, and rollback checks for GitHub Copilot Enters Slack Public Preview for Shared Coding Work.
Summary
Memory and Ollama make JetBrains workflows more continuous and flexible. Start with non-sensitive tasks, prove the data path, and expand repository access only after the evidence is clear.
