安全与隐私

AI agent engineering readings for September 11 to 17, 2026, with key findings, engineering implications, suggested sections, and primary sources.

AI agent engineering readings for September 11 to 16, 2026, with key findings, engineering implications, suggested sections, and primary sources.

How the Google Cloud Developer Plugin packages Skills, MCP configuration, and rules, and what this means for distribution, versions, and permissions.

OpenCompany uses company.toml, agent teams, and human sign-off to put organizational structure and agent configuration into version-controlled files.

The Anthropic incident review shows why automated monitors and LLM graders need coverage, independence, and tests for their own failure modes.

AI agent engineering readings for September 4 to 10, 2026, with key findings, engineering implications, suggested sections, and primary sources.

Coding agents read and execute skills, MCP tools, dependencies, and repository instructions. These need software supply-chain controls.

How GitHub Agentic Workflows puts judgment-based automation inside GitHub Actions while retaining permissions, controlled outputs, approval, and security.

How sandboxes limit untrusted code access to files, processes, networks, and resources, comparing VMs, containers, user-space kernels, and syscall filters.