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.
How the EU AI Act and NIST AI RMF apply to agents that rewrite their own skills and memory, with dates, article text, a control map and setup steps.
Persistent prompt injection happens when an agent saves injected text as memory or a skill. Learn the paths, the research, and the controls that undo it.
Inside Hermes Agent: six entry points feed one AIAgent loop with a tool registry, sandboxed terminal backends, a messaging gateway and an MCP client.
Gate an agent's own memory, skill and config writes by reversibility, log every change outside its reach, test it on a canary, and keep a tested rollback.
DSPy, OPRO, Promptbreeder and TextGrad rewrite prompts and examples against a metric. How each works, what the papers showed, and how to govern the output.
Self-training LLMs learn from their own filtered outputs. How STaR, Self-Rewarding models and self-play work, what they showed, and where they break.
Install Hermes Agent from a pinned stable release, move commands into Docker, keep keys out, default-deny egress and stage every skill and memory write.
DGM, ADAS, Gödel Agent and SICA let an LLM agent rewrite its own code and keep changes that score better. What each changed, cost and risked.
Continual learning means a deployed model or agent keeps changing. How feedback loops, forgetting and drift break it, and how to monitor and roll it back.
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.
