Note
SWRMZ is currently in beta. Request early access to get on the platform when we launch.
SWRMZ is a swarm of specialized AI security agents that continuously find vulnerabilities, fix them automatically, report every action, and guard your logs around the clock. Built on Band.ai. One swarm, zero blind spots.
The problem: Security tooling floods you with alerts, most of which are noise. Developers tune them out. Vulnerabilities slip through. Breaches happen.
Our solution: A coordinated swarm of agents that understands your stack contextually, validates real threats through exploitation, opens fix PRs automatically, and keeps watch — before attackers find what you missed.
Point the swarm at your repos, cloud, and log streams and it works nonstop:
- Deep analysis across code, dependencies, and infrastructure
- Validated findings — recon, analysis, then real exploitation, so no exploit means no report
- Zero configuration — agents map your attack surface in minutes
Detection without remediation is just a to-do list. SWRMZ fixes what it finds:
- Auto-generated PRs with safe, reviewable code changes
- Contextual fixes that respect your codebase patterns
- Approval policy — auto-apply what you trust, review the rest
Every finding, exploit proof, and fix is documented for you:
- SARIF output and executive summaries
- Remediation plans ready to hand to auditors
- Full history of what changed and why
Security isn't a one-time scan. The swarm runs as long as your stack evolves:
- Real-time log watching with anomaly detection
- Threat lockdown before issues spread
- Drift detection as your codebase grows
1. Deploy → Point SWRMZ at your repos, cloud, and log streams
2. Recon → Agents map your attack surface in minutes
3. Detect → Real vulnerabilities surfaced, noise filtered
4. Fix → Automated PRs with secure replacements
5. Guard → Continuous monitoring, reporting, and lockdown
SWRMZ uses Band as the real coordination layer for the desktop scan flow, not as a final notification channel.
- The Electron app authenticates as a Band Operator and creates a fresh Band room for each scan.
- The Operator recruits three registered Band remote agents:
SWRMZ Analyst,SWRMZ Fixer, andSWRMZ Reviewer. - The scan request is posted into that Band room. Analyst hands findings to Fixer, Fixer hands patch context to Reviewer, and Reviewer signs off through Band mentions.
- The app shows the live Band room id and mirrors the room transcript beside the local scan/fix engine so the handoffs are visible to the user and auditable in the saved session.
Relevant source: app/electron/band_bridge.cjs, app/electron/main.cjs, and band_agents/.
Local proof command, after configuring the four Band keys and starting band_agents/run.py:
node app/electron/smoke_band_flow.cjsIt creates a temporary vulnerable repo, opens a real Band room, waits for Operator → Analyst → Fixer → Reviewer messages, and prints the room id plus sender summary.
| Layer | Technology |
|---|---|
| Framework | React 19, TypeScript, Vite |
| Routing | TanStack Router |
| Styling | Tailwind CSS v4 |
| Animation | Framer Motion |
| Icons | Lucide |
| AI Engine | Band.ai (separate service) |
| Hosting | Vercel |
├── public/ # Static assets (WebP imagery, favicons)
├── src/
│ ├── components/ # Shared UI (BrandMark, AppShowcase, StepArt)
│ ├── sections/ # Landing sections
│ │ ├── HeroSection.tsx
│ │ ├── PlatformSection.tsx
│ │ ├── WorkflowSection.tsx
│ │ ├── TestimonialsSection.tsx
│ │ ├── PricingSection.tsx
│ │ ├── FaqSection.tsx
│ │ └── FinalCta.tsx
│ ├── LandingPage.tsx # Marketing page composition
│ ├── LoginPage.tsx # /login — split sign-in screen
│ ├── main.tsx # App entry + router
│ └── styles.css # Tailwind + keyframes
└── vercel.json # Deployment config (SPA rewrites)
- Clone & Install:
git clone https://github.com/YUST777/swrmz.git
cd swrmz
npm install- Run Dev Server:
npm run dev- Build for Production:
npm run build
npm run previewMarketing site, interactive scan-session showcase, pricing, FAQ, and the sign-in surface. Foundation is live.
User authentication (email, SSO, passkey), organization management, and the swarm dashboard.
The core product — autonomous detection, remediation, reporting, and log guarding with repo and cloud integration.
GitHub Actions, GitLab CI, and IDE plugins for seamless developer-workflow integration.
- Bug Reports: Open an Issue
- Feature Requests: Request a Feature
- Early Access: Get started at swrmz.ai
Give us a star on GitHub — it helps us grow and keeps the momentum going.