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.
PromptArmor showed in 2024 that Slack AI could be steered by a public channel message to leak a private key. How it worked and how scope limits help.
A decision guide for letting a coding agent run npm or pip installs on its own: when approval is wise, what to restrict, and why install scripts matter.
One design idea for injection resistant agents: keep untrusted text away from the model that plans actions. What the pattern is and what is still research.
A rules file steers an assistant toward safer defaults. What to include, what it cannot guarantee, and why the file needs the same review as code.
How to scope an AI red team exercise in three phases, plan, execute, then assess and report, following the MITRE ATLAS AI Red Team mitigation.
Secrets, personal data, hidden text and unreviewed sources can all end up in your search index. A pre-indexing checklist you can run as a script.
A local MCP server runs with your privileges. Read the exact startup command, restrict files and network, and prefer stdio. A practical sandboxing guide.
In 2023 Samsung restricted generative AI after staff uploaded source code to ChatGPT. What the memo said, what remains unknown, and the lessons.
Safetensors stores tensors without executable code, which removes the pickle risk. It does not prove a model is trustworthy. Here is the difference.
Weekly insights on software supply chain security, delivered to your inbox.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.
