feat: require dedicated security section in all reviews

[LearningCircuit]

Summary

• Adds mandatory 🔒 Security section to all AI code reviews
• Ensures security is always explicitly evaluated
• If no concerns found, reviewer must state "No security concerns identified"

Changes

Updated the review structure in ai-reviewer.sh to require:

1. Short overall feedback summary (1-2 sentences)
2. 🔒 Security section (new requirement)
3. Other detailed findings
4. Final verdict

Benefits

• Makes security review explicit and consistent across all PRs
• Prevents security concerns from being overlooked
• Clear "all clear" message when no issues exist
• Keeps reviews focused on high-value security analysis

Test plan

• Test with a PR containing potential security issues
• Verify security section is always present
• Verify "No security concerns identified" appears when appropriate

[github-actions]

AI Code Review

Clean implementation of a valuable process improvement that makes security review explicit and consistent across all PRs.

🔒 Security

• No security concerns identified in this change

Code Quality & Best Practices

• Clear and unambiguous: Explicit requirement with clear "no concerns" messaging prevents ambiguity
• Correct shell escaping: " properly escaped for string context in ai-reviewer.sh
• Consistent formatting: Maintains existing numbered structure and verdict options
• Process improvement: Ensures security is never accidentally omitted from reviews

Minor Consideration

Inference (not verified): If automated validation of review format exists, confirm it's updated to enforce the security section requirement.

✅ Approved with recommendations

---

Review by Friendly AI Reviewer - made with ❤️