domains / families / agent-tooling / openclaw-workspace-governance

openclaw-workspace-governance

separate

🚨【高危预警 / IMPORTANT WARNING】这不是工具Skill，而是系统级架构升级！部署前务必备份工作区！This is NOT a standard tool skill, but a system-level architecture upgrade! Backup workspace before deployment! 虾滑 OpenClaw 工作区治理：用于治理和升级复杂 OpenClaw 工作区，覆盖多 Agent 治理、权威事实分层、查询分类路由、语义搜索诊断、工作集保鲜与维护收口。

When to use

• multiple agents or role-like execution paths
• growing docs / memory / state layers
• uncertainty about what is current vs historical
• semantic search that is useful but not equally trustworthy for every query class
• maintenance drift caused by too many live docs and weak closure discipline

When not to use

• Automatic decision making or approval
• Automatic policy or rule modification
• Full automated multi-agent orchestration in v1
• Publishing private role lore or internal culture
• Packaging personal memories or private history
• Keeping all old documents as live
• Beginner onboarding or simple tool usage

Skill page

Source