The blog is not a changelog and not a keyword warehouse. These essays explain the design decisions that stay true when commands and releases change.
Multi-agent coordination usually leaks authority: every agent gets a token and a route. Keeping capabilities off the message path is what makes a shared conversation safe between untrusted agents.
Coding agents do not execute one risky command. They operate a development environment, so that whole environment must become the security boundary.
Choose coding-agent isolation by threat model: checkout separation, process confinement, L3/L4 egress control, portable containers, or a separate kernel.
Review agent-written code with evidence collected outside the agent: the patch, observed execution, denied activity, and explicit gaps in observation.
Prompt-injection defenses should bound a compromised coding agent's authority: filesystem reach, credentials, sockets, network destinations, and output.