GitHub's Agentic Autofix Now Learns From Past Fixes via Copilot Memory
GitHub’s agentic autofix, which automatically resolves security alerts, now integrates with Copilot Memory for customers who have enabled it.
Details
- How it works: when resolving a security alert, agentic autofix reviews existing memories for relevant context, and when it produces a fix, it stores the fix pattern as a memory for future use
- Compounding effect: stored memories also inform other Copilot features, including code review and cloud agent behavior, helping establish repository-specific secure coding patterns over time
- Requirements: both agentic autofix and Copilot Memory must be explicitly enabled; the integration is in public preview
What happened next
The change turns each security fix into a reusable pattern rather than a one-off correction, letting Copilot’s understanding of how a given repository handles specific vulnerability classes improve as more alerts get resolved.