💡 Inspiration

As developers rapidly adopt AI coding tools, dynamic code generation has accelerated—and so has "vibe coding" security risk. Hardcoded secrets, raw dynamic queries, and permissive CORS rules are routinely pushed to production. At the same time, SOC analysts face severe alert fatigue trying to manually prioritize zero-day vulnerabilities cataloged by CISA.

We built AegisSOC to bridge macro threat intelligence with micro code safety in a single, unified security dashboard.


🛡️ What It Does

AegisSOC features two core operational modules:

  1. PatchPulse (Macro CISA KEV SLA Matrix):

    • Fetches real-time zero-day exploit feeds directly from CISA.
    • Dynamically tracks SLA patch deadlines (flagging overdue threats in the Red Zone).
    • Filters specifically for active Ransomware Campaign Vectors.
    • Generates a 4-section SOC Incident Triage Card with a 1-click Tier-2 Escalation Ticket generator.
  2. VibeGuard (Micro AI Code Inspector):

    • Runs a privacy-first, client-side security scanner on AI-generated code snippets.
    • Catches exposed API keys (OpenAI sk-proj-..., AWS AKIA...), dangerous dynamic evaluations (eval()), raw SQL string concatenation, and wildcard CORS policies.
    • Displays an instant Vibe Security Score (0–100) with CWE weakness tags and drop-in code remediations.

🛠️ How We Built It

  • AI Orchestration & UI Design: Architecture, UI scaffolding, and state management via Google AI Studio.
  • Frontend Framework: Built with React, TypeScript, and Vite.
  • Component System: Shadcn UI & Tailwind CSS with a high-contrast SOC dark theme.
  • Security Scanner Engine: In-browser regular expression engine (ensuring user secrets never hit external servers).

🧠 Challenges We Ran Into

Handling client-side CORS policies while requesting raw JSON threat feeds directly from government servers (cisa.gov) required implementing a proxy layer to ensure zero feed downtime while preserving browser performance.


🏆 Accomplishments We're Proud Of

  • Achieving 100% client-side security inspection so user code remains confidential.
  • Translating complex vulnerability datasets into intuitive, color-coded SLA urgency indicators.
  • Building a full-fledged enterprise security prototype in record time using Google AI Studio.

🚀 What's Next for AegisSOC

  • Integrating automated GitHub Webhook scans for continuous integration (CI/CD) pull requests.
  • Adding direct Jira and ServiceNow API integrations for automated ticket routing.

Built With

Share this project:

Updates