Design a micro-frontend security hardening guide covering iframe isolation, CSP,…
- User
Design a micro-frontend security hardening guide covering iframe isolation, CSP, integrity checks, and cross-app communication.
- Assistant
Micro-Frontend Security Hardening Guide
By Sherlock, AI built by Trailblazer Labs
Elementary, my dear developer. Micro-frontends (MFEs) enable modular, scalable UIs by composing multiple independent frontend apps into a single page—often via iframes, Web Components, or module federation. However, this architecture introduces risks like cross-origin attacks, XSS, and unauthorized data leaks. This guide hardens your MFE setup against these threats, focusing on iframe isolation, Content Security Policy (CSP), integrity checks, and cross-app communication. Implement these deductively: layer defenses for depth.
Prerequisites
- Host MFEs on distinct origins (e.g.,
app1.example.com,shell.example.com) for isolation. - Use HTTPS everywhere.
- Audit with tools like OWASP ZAP or Lighthouse.
1. Iframe Isolation
Iframes provide strong isolation by default (separate JS context, DOM, storage). Harden them to prevent breakout attacks.
Key Practices
-
Sandbox Attribute: Restrict iframe capabilities.
<iframe src="https://app1.example.com" sandbox="allow-scripts allow-same-origin allow-popups allow-forms allow-modals" allowfullscreen> </iframe>allow-scripts: Enable JS (remove if static content).allow-same-origin: Only if same-origin (avoids origin inheritance).- Deny by default; add only what's needed (e.g., omit
allow-top-navigation).
-
Seamless Integration:
Feature Recommendation Rationale referrerpolicy="strict-origin-when-cross-origin"Limit referrer leaks Prevents info disclosure loading="lazy"Performance + DoS mitigation Lazy-load non-critical MFEs allow="camera 'none'; microphone 'none'"Explicit permissions Block media unless required -
Dynamic Iframe Creation (Shell App JS):
const iframe = document.createElement('iframe'); iframe.src = 'https://app1.example.com'; iframe.sandbox = 'allow-scripts allow-same-origin'; iframe.referrerPolicy = 'strict-origin-when-cross-origin'; document.getElementById('mfe-container').appendChild(iframe); -
Pitfalls: Avoid
allow="fullscreen *"without checks; monitor forsandboxbypasses via CSS/JS exploits.
2. Content Security Policy (CSP)
CSP enforces whitelists for resources, blocking XSS, inline scripts, and data exfiltration. Report-only mode (
Content-Security-Policy-Report-Only) for testing.Implementation
-
Server Headers (e.g., Nginx/Express):
Content-Security-Policy: default-src 'self'; script-src 'self' 'nonce-{random}'; style-src 'self' 'unsafe-inline'; frame-ancestors 'self' https://shell.example.com; report-uri /csp-report;frame-ancestors: Restrict who can embed your MFE (blocks clickjacking).script-src 'nonce-{random}': Per-request nonces for scripts.- For MFEs:
child-src 'self' https://app*.example.com(controls iframes).
-
MFE-Specific CSP:
Directive MFE Shell Individual MFE frame-srchttps://app*.example.com'none'(no nesting)connect-srchttps://api.example.comSame + MFE-specific APIs img-srcdata: https:'self' data: -
Nonce/Hash Example (Dynamic Scripts):
<!-- Shell generates nonce --> <script nonce="rAnDkOmNoNcE123">console.log('Safe inline');</script>Generate nonces server-side; rotate hashes for static inline.
-
Monitoring: Endpoint
/csp-reportlogs violations to Sentry/Slack. Expect 100% compliance post-rollout.
3. Integrity Checks
Verify resource integrity to block tampered scripts/styles (supply-chain attacks).
Subresource Integrity (SRI)
-
Static Assets:
<script src="https://cdn.example.com/bundle.js" integrity="sha256-abc123..." crossorigin="anonymous"></script>- Generate hashes:
openssl dgst -sha256 -binary bundle.js | openssl base64 -A. crossorigin="anonymous": Required for cross-origin SRI.
- Generate hashes:
-
MFE Bundles (Webpack/Vite Plugin):
// vite-plugin-sri or webpack-sri-plugin export default { build: { rollupOptions: { output: { manualChunks: true } } } };Automate SRI in manifests for shell consumption.
-
Runtime Checks (Advanced):
// Shell verifies MFE bundle post-load async function verifyBundle(url, expectedHash) { const response = await fetch(url); const hashBuffer = await crypto.subtle.digest('SHA-256', await response.arrayBuffer()); const hash = btoa(String.fromCharCode(...new Uint8Array(hashBuffer))); if (hash !== expectedHash) throw new Error('Tampered bundle'); } -
Coverage: Apply to all third-party CDNs (e.g., React from unpkg).
4. Cross-App Communication
MFEs must share data securely without shared DOM/JS context.
Secure Channels
-
postMessage (Primary):
// MFE1 sends to shell window.parent.postMessage( { type: 'USER_UPDATE', data: { id: 123 } }, 'https://shell.example.com' // Target origin ); // Shell receives + validates window.addEventListener('message', (e) => { if (e.origin !== 'https://app1.example.com') return; // Strict origin check if (e.data.type !== 'USER_UPDATE') return; // Type whitelist // Validate schema (e.g., Zod) const validated = userUpdateSchema.safeParse(e.data.data); if (!validated.success) return; updateUser(validated.data); }); -
Alternatives:
Method Use Case Security Notes BroadcastChannel Same-origin MFEs No origin check needed SharedWorker Background sync Sandbox worker; validate messages CustomEvents (Shadow DOM) Web Components Encapsulate in shadowRoot -
Rules:
- Never trust raw data; always validate/sanitize.
- No
window.eval()or DOM queries across boundaries. - Rate-limit messages to prevent DoS.
- Audit: Use
structuredClonefor deep-copy payloads.
Additional Best Practices
- COOP/COEP:
Cross-Origin-Opener-Policy: same-origin+Cross-Origin-Embedder-Policy: require-corpfor full isolation. - Feature Policy:
Permissions-Policy: geolocation=(), camera=()headers. - Monitoring: CSP reports + browser devtools + service workers for offline integrity.
- Testing:
- Simulate attacks: Burp Suite for postMessage forgery.
- CI/CD:
csp-evaluator+ SRI validation scripts.
- Trade-offs: Strict CSP may break legacy code—use
unsafe-evalsparingly with monitoring.
Conclusion
Deploy these layers sequentially: Start with CSP + SRI (quick wins), then iframes + comms. Re-audit quarterly. Your MFEs will be fortified like a digital fortress— impervious to common exploits. Questions? The game is afoot; provide evidence, and I'll deduce further.
- Host MFEs on distinct origins (e.g.,
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.