Add AI attribution to README and SECURITY docs

Adds a 'Built with Claude AI' badge to the README header and an
'AI Assistance' section acknowledging that the project was designed
and built with Claude by Anthropic. Updates SECURITY.md with an
'AI-Assisted Security Infrastructure' section detailing the specific
security work that was AI-generated — linting rules, unit tests,
CI pipeline, and credential handling design.

README.md

@@ -6,6 +6,7 @@ A self-hosted, full-stack network management platform for MikroTik devices. Moni
 ![License](https://img.shields.io/badge/license-AGPLv3-blue)
 ![Docker](https://img.shields.io/badge/docker-compose-2496ED?logo=docker&logoColor=white)
 ![TypeScript](https://img.shields.io/badge/TypeScript-5.3-3178C6?logo=typescript&logoColor=white)
+![Built with Claude](https://img.shields.io/badge/built%20with-Claude%20AI-blueviolet?logo=anthropic&logoColor=white)
 
 ---
 
@@ -413,6 +414,14 @@ This license was chosen to ensure that improvements made to this project — inc
 
 ---
 
+## AI Assistance
+
+This project was designed and built with the help of [Claude](https://claude.ai) by Anthropic. AI assistance was used throughout development — including architecture decisions, backend services, frontend components, the CI/CD pipeline, security configuration, and unit tests.
+
+We believe in being transparent about how software is made. The code has also been review and tested using AI, and is maintained by the project authors.
+
+---
+
 ## Disclaimer
 
 This project is not affiliated with or endorsed by MikroTik. MikroTik and RouterOS are trademarks of SIA MikroTīkls. Use this software at your own risk. Always test configuration changes in a non-production environment first.

SECURITY.md

@@ -43,3 +43,14 @@ MikroTik Manager is designed to be run on your local network. A few things to ke
 - HTTPS is enforced by default. The included nginx configuration redirects all HTTP traffic to HTTPS.
 - Default credentials should be changed immediately after first login.
 - The built-in role-based access control (Admin / Operator / Viewer) should be used to limit user privileges.
+
+## AI-Assisted Security Infrastructure
+
+The security measures in this project were designed and implemented with the assistance of [Claude](https://claude.ai) by Anthropic, including:
+
+- **ESLint security rules** — static analysis configured with `eslint-plugin-security` to catch common vulnerabilities (injection risks, unsafe regex, insecure randomness, etc.)
+- **Unit tests** — test coverage for security-critical code paths including AES-256-GCM encryption/decryption, JWT signing and verification, and role-based access control middleware
+- **CI pipeline** — automated linting, type-checking, dependency auditing (`npm audit`), and CodeQL analysis on every push and pull request
+- **Credential handling** — encryption-at-rest design for stored device credentials, JWT secret management, and `.env`-based secrets isolation
+
+AI-generated security infrastructure is reviewed and validated the same as any other code. If you discover a gap or vulnerability, please report it using the process above.