Developer Tools
Maven mirrorOf: wildcards, exclusions and the whitespace trap
How Maven mirrorOf values like star, external:star and star,!repo work, how order decides which mirror wins, and why a space can mirror nothing.
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 feedDeveloper Tools
Maven BOMs and dependencyManagement: keeping versions in one place
How dependencyManagement and a BOM import centralise Maven versions, what the import scope does, and how declaration order settles conflicts.
Oct 18, 20263 min read
Developer Tools
Matter device attestation: why a fake device cannot just join
Matter requires devices to prove they are genuine before joining a fabric. How attestation works, what the ledger does, and what it does not cover.
Oct 18, 20264 min read
Developer Tools
Match patterns explained: how extensions get access to sites
Match patterns decide which sites a browser extension can touch. Learn to read them, spot the risky ones like all_urls, and judge a request quickly.
Oct 18, 20263 min read
Developer Tools
Mass assignment: DTOs, allowlists or blocklists?
Binding request data onto models lets attackers set fields like isAdmin. Compare DTOs, field allowlists and blocklists, and why authorization matters.
Oct 18, 20263 min read
Developer Tools
Maintainer boundaries: what to document in your README
Open Source Guides suggests writing down what you will and will not do. Here is a README section that sets expectations and protects a maintainer's time.
Oct 18, 20263 min read
Developer Tools
Lost in the middle: why stuffing more context into a model can backfire
Research shows models use information at the start and end of long inputs best. What that means for how many chunks you retrieve and where you place them.
Oct 18, 20263 min read
Developer Tools
Long context pricing: why a huge prompt can change your rate
Some AI models charge more per token once a prompt passes a size threshold. Dated examples from Anthropic, Google and OpenAI, and what to do about them.
Oct 18, 20263 min read
Developer Tools
Long inputs in security prompts: put code first, question last
How to order a prompt that includes many files or a long log: documents at the top, labeled by source, a quote step, then your question at the end.
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.