MCP 与 Skills

Seven updates on Codex prewarming and Guardian, remote sandboxes, MCP OAuth, secret stores, and agent software factory production data.

7 条 Agent 工程更新，覆盖 Codex 会话预热和 Guardian、远程 Sandbox、MCP OAuth、Secret Store 与 Agent 软件工厂数据。

5 条 Agent 工程更新，重点看 Tool 与 Skill 上下文预算、context doctor、持久调度、Tool Loop 预算和 fail-closed 审批。

Five updates on tool and skill context budgets, context doctor, persistent scheduling, tool-loop limits, and fail-closed approvals.

Seven updates on token efficiency, Codex networking, Copilot Memory, local agent server security, self-practice, skills, and reproducibility.

7 条 AI Agent 工程更新，覆盖长任务 Token 优化、Codex 网络策略、Copilot Memory、本地 Agent Server 安全和 Self-Practice。

Five updates on proactive safety intervention timing, state revision, TeamHarness, enterprise MCP permissions, and tool-result handling.

5 条 AI Agent 工程更新，重点看主动安全介入时机、状态修正、TeamHarness、企业 MCP 权限边界和 Tool 结果处理。

Plugin4Shell exposes a supply-chain gap: a pinned SHA records the expected version, but installers must also verify the actual HEAD after checkout.

Plugin4Shell 暴露了一个供应链边界：固定 SHA 只写下期望版本，安装器还必须在 checkout 后验证真实 HEAD。本文结合 Git 文档、本地无害复现和 GitHub 分支限制解释机制。

LangGraph human-approval schemas, Deep Agents safeguards for parallel writes to the same file, and tighter OAuth boundaries in the MCP Ruby SDK.

本期收录 LangGraph 人工审批结构约束、Deep Agents 同文件并行写保护，以及 MCP Ruby SDK OAuth 边界收紧。