AI Security
Auditing AI agent skill registries for hardcoded keys
29M new hardcoded secrets hit public GitHub in 2025, up 34% YoY — and 3% of MCP servers in production carry hardcoded credentials as theft traps.
Jul 9, 20267 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 feedAI Security
Scanning AI-Generated Code Before It Merges: Wiring Scanners into Coding Assistants with MCP
Research found ~40% of Copilot suggestions were vulnerable, and devs using AI assistants trusted their code more. MCP lets you scan before merge.
Jul 9, 20267 min read
Vulnerability Management
SBOM-based blast radius analysis for vulnerable dependencies
An SBOM tells you what's inside one artifact. It takes a dependency graph across every service to know what breaks first when a library gets a CVE.
Jul 9, 20266 min read
Application Security
Reducing false positives in SAST and SCA tools
NIST benchmark data puts some SAST false-positive rates near 78%. Reachability analysis and contextual triage are how teams cut that noise without missing real risk.
Jul 9, 20266 min read
Application Security
Inside CVE-2025-55182: the React Server Components RCE and how to defend against it
A CVSS 10.0 pre-auth RCE in React Server Components, exploited within 48 hours of disclosure — how the deserialization flaw works and how to mitigate it.
Jul 9, 20266 min read
AI Security
A Reproducible Rubric for Measuring Prompt-Injection Risk in Agent Skills
OWASP has ranked prompt injection the #1 LLM risk for two straight editions, yet almost no one scores agent skill packages for it consistently. Here's a rubric.
Jul 9, 20266 min read
Open Source Security
The 4 dimensions of open-source dependency risk
Open-source risk isn't one problem — CVEs, malware, license exposure, and abandonment each fail differently, and Sonatype logged 454,600+ malicious packages in 2025 alone.
Jul 9, 20267 min read
Supply Chain Attacks
Inside the npm Reward-Farming Worm That Published 89,000+ Packages
One npm publishing bot exploited a crypto reward protocol to spam 89,000+ packages, some appearing every 7-10 seconds. Here's how it worked and how to spot it.
Jul 9, 20267 min read
Supply Chain Attacks
The npm worm incident response playbook
Shai-Hulud compromised 500+ npm packages by auto-publishing itself with stolen tokens. Here's a concrete detection, rotation, and pinning playbook.
Jul 9, 20265 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.