Security Guides
How to read the systemd-analyze security exposure score
systemd-analyze security rates a service from 0.0 to 10.0. What the number means, what it cannot see, and how to use it to pick the next setting to add.
Oct 18, 20263 min read
Resources
Supply Chain Security, in plain English.
Deep dives, practical guides, and incident analyses from engineers who build Safeguard. No fluff, no vendor FUD — just what you need to ship secure software.
Filter
Articles
RSS feedSecurity Guides
React Native deep links and OAuth: why PKCE is required
Custom URL scheme deep links can be hijacked by another app. Learn why React Native sign in flows need PKCE and what never to put inside a link.
Oct 18, 20263 min read
Security Guides
Is a vulnerable Java dependency actually reachable in my app?
Present means the library is on your classpath. Reachable means your code can call the vulnerable function. How to tell them apart in Java and what to fix.
Oct 18, 20263 min read
Security Guides
Reachable or exploitable: which question should you ask first?
Reachability asks whether your code calls the vulnerable function. Exploitability asks whether an attacker can use it. Ask the first, then the second.
Oct 18, 20263 min read
Security Guides
Does reachability analysis work across internal packages in a monorepo?
A vulnerable function is reached through your own packages too. Learn why call paths must cross internal package boundaries and what to ask a tool.
Oct 18, 20263 min read
Security Guides
RBI VA and PT schedule: every 6 and 12 months for critical systems
RBI sets vulnerability assessment every six months and penetration testing every 12 months for critical systems. What else section 26 asks of teams.
Oct 18, 20263 min read
Security Guides
Which RBAC roles should an enterprise AppSec tool support?
A practical role model for AppSec tools: viewer, triager, policy admin, auditor and billing, plus the scoping and test questions to ask during a pilot.
Oct 18, 20264 min read
Security Guides
Should you score threats with numbers or High, Medium, Low?
OWASP prefers qualitative High, Medium, Low ratings over numeric threat scores. How to rate by ease of exploitation and impact, with a worked table.
Oct 18, 20264 min read
Security Guides
When the rate limiter's datastore dies: fail open or closed?
A rate limiter backed by a shared store must decide what to do when the store is gone. A route by route guide to fail open, fail closed or local fallback.
Oct 18, 20263 min read
Stay informed
Weekly insights on software supply chain security, delivered to your inbox.
Self-healing security runs on Safeguard.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.