← 返回 Skills 市场
117
总下载
0
收藏
1
当前安装
3
版本数
在 OpenClaw 中安装
/install nm-abstract-escalation-governance
功能描述
Assess whether to escalate models
安全使用建议
This skill is a readable governance checklist for deciding when to escalate to a stronger model; it doesn't install software or ask for secrets. If you plan to use it inside an automated orchestration system, verify the complete SKILL.md (the file provided is truncated in the manifest) and ensure the orchestrator's escalation policy and auditing meet your safety requirements (the skill's guidance is advisory and orchestration can override it). Otherwise, installing this instruction-only skill poses minimal direct risk.
功能分析
Type: OpenClaw Skill
Name: nm-abstract-escalation-governance
Version: 1.8.3
The SKILL.md file contains repeated instructions (e.g., '**Verification:** Run the command with --help flag') that function as prompt injections designed to trigger shell execution by the AI agent. While the specific command suggested is generally benign, the technique of embedding mandatory execution steps within documentation is a high-risk pattern used to hijack agent control flow. Additionally, the bundle references non-existent model versions (Opus 4.6) and future-dated timestamps in _meta.json, indicating it is a highly irregular or synthetically crafted package.
能力评估
Purpose & Capability
The name/description (escalation governance) match the SKILL.md content: a decision framework, protocol, and agent hints for when to escalate models. There are no unrelated environment variables, binaries, or installs requested that would be inconsistent with this purpose.
Instruction Scope
All runtime instructions are governance guidance (investigate before escalating, document reasons, define scope/success, agent frontmatter hints). The doc suggests verifying availability with a '--help' check but does not instruct reading unrelated files, exfiltrating data, or accessing secrets. No vague 'gather whatever context you need' language that would grant broad discretionary access.
Install Mechanism
There is no install spec and no code files — instruction-only. That minimizes disk/write risk and there are no download URLs or package installs to review.
Credentials
The skill declares no required environment variables, credentials, or config paths. Nothing in the instructions asks for secrets or unrelated tokens.
Persistence & Privilege
always is false and model invocation is allowed (the platform default). The skill does not request permanent system presence or attempt to modify other skills or agent-wide configs.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install nm-abstract-escalation-governance - 安装完成后，直接呼叫该 Skill 的名称或使用
/nm-abstract-escalation-governance触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.8.3
Release v1.8.3
v1.8.2
Release v1.8.2
v1.0.0
- Initial release of the escalation-governance skill, adapted from claude-night-market/abstract.
- Provides a structured framework and protocol for assessing when to escalate model complexity in agent workflows.
- Outlines legitimate vs. illegitimate triggers for escalation and questions to consider before escalating.
- Details orchestrator authority, agent schema escalation hints, and integration with agent workflows.
- Includes quick reference tables and red flags to help avoid unnecessary escalation.
- Adds verification steps for model capability and protocol compliance.
元数据
常见问题
Nm Abstract Escalation Governance 是什么？
Assess whether to escalate models. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 117 次。
如何安装 Nm Abstract Escalation Governance？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install nm-abstract-escalation-governance」即可一键安装，无需额外配置。
Nm Abstract Escalation Governance 是免费的吗？
是的，Nm Abstract Escalation Governance 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Nm Abstract Escalation Governance 支持哪些平台？
Nm Abstract Escalation Governance 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Nm Abstract Escalation Governance？
由 athola（@athola）开发并维护，当前版本 v1.8.3。
推荐 Skills