On branch edburns/re-enable-pre-commit-hooks

modified:   .github/copilot-instructions.md

Apply @Copilot review recommendation.

> This new section says the pre-commit hook is “automatically enabled” via `copilot-setup-steps.yml`, but earlier in the same document (and in CONTRIBUTING/README) the hook is described as something developers must manually enable with `git config core.hooksPath .githooks`. Please clarify the scope (e.g., auto-enabled only in the Copilot coding agent environment; local development still requires manual enablement) to avoid conflicting guidance, and consider linking explicitly to `.github/workflows/copilot-setup-steps.yml` for precision.

Signed-off-by: Ed Burns <edburns@microsoft.com>

Author: edburns

.github/copilot-instructions.md

@@ -104,7 +104,7 @@ When porting from .NET:
 - 4-space indentation (enforced by Spotless with Eclipse formatter)
 - Fluent setter pattern for configuration classes (e.g., `new SessionConfig().setModel("gpt-5").setTools(tools)`)
 - Public APIs require Javadoc (enforced by Checkstyle, except `json` and `events` packages)
-- Pre-commit hook runs `mvn spotless:check` - enable with: `git config core.hooksPath .githooks`
+- Pre-commit hook runs `mvn spotless:check` - Must be manually enabled with: `git config core.hooksPath .githooks`, except in the Copilot coding agent environment. This hook is explicitly enabled in the Copilot coding agent environment. See [copilot-setup-steps.yml](workflows/copilot-setup-steps.yml).
 
 ### Handler Pattern