Skip to content
All projects
Application Security / SSO & SAMLCompletedShipped Jun 2026

SAMLCheck — SSO Metadata Validator

A security-first web app that validates SAML metadata XML in seconds — extracting entity IDs, ACS/SLO endpoints, and X.509 certificates with PASS/WARN/FAIL verdicts and a REST API, processed entirely in memory.

Proposed stack

  • TypeScript
  • Node.js / Express
  • Vite
  • fast-xml-parser
  • REST API

Goal

Give identity and security teams a fast, safe way to validate SAML/SSO metadata without exposing sensitive XML.

Overview

SAMLCheck ingests a SAML metadata XML file (upload or paste) and validates it instantly — extracting the Entity ID / Issuer, Assertion Consumer Service (ACS) URLs and bindings, Single Logout endpoints, and X.509 certificates, including expiration checks.

Security is the core design constraint: files are processed entirely in memory and never written to disk or logged, and the XML is parsed with fast-xml-parser under strict settings that reject external entities to prevent XXE and entity-expansion attacks.

Results come back as a clear PASS / WARN / FAIL verdict with categorized findings, and the same validation is available programmatically through a REST /api/validate endpoint.

Architecture

SAML Metadata XML → Secure Parser → Validation Engine → Verdict → Report & API

All validation happens in memory — raw XML is never stored or logged. The same engine backs both the UI and a REST /api/validate endpoint.

Key features

  • Upload or paste SAML metadata XML (drag-and-drop supported).
  • Extract Entity ID / Issuer, ACS URLs & bindings, and Single Logout endpoints.
  • Validate X.509 certificates, including expiration.
  • PASS / WARN / FAIL verdicts with categorized findings.
  • In-memory-only processing — no storage, no logging of raw XML.
  • Secure XML parsing hardened against XXE and entity-expansion attacks.
  • REST /api/validate endpoint for programmatic use.

Technical challenges

  • Parsing untrusted XML safely — hardening against XXE and entity-expansion attacks while still extracting everything needed.
  • Deciding what should PASS, WARN, or FAIL so results are genuinely actionable for identity and security teams.
  • Keeping sensitive metadata in memory only, with no storage or logging.

How it's built

  • 1Express + Vite (TypeScript) app with a shared validation engine behind both the UI and the API.
  • 2fast-xml-parser configured to reject external entities; XML cleared from memory immediately after validation.
  • 3Validation checks for Entity ID, ACS/SLO endpoints, and X.509 certificate expiration, producing categorized findings.
  • 4RESTful /api/validate endpoint with JSON input/output for programmatic integration.

Interested in this work?

Happy to talk through the design, trade-offs, and implementation plan.

Get in touch