Nm Leyline Utility

Score candidate agent actions by expected gain, cost, uncertainty, and redundancy to guide dispatch and termination decisions

This skill is internally coherent and low-risk: it only provides heuristics and requires no credentials or external installs. Before enabling it in production, confirm (1) which consuming skills will be allowed to set utility_gated:true (prescriptive mode can force actions), (2) that you are comfortable granting consumer skills access to conversation history/TaskList/token counters, and (3) that consumers correctly log/justify any overrides. Absence of regex findings is expected for an instruction-only skill — review integrations and frontmatter of consumer skills that adopt it to ensure they don't inadvertently grant excessive privileges.

Type: OpenClaw Skill Name: nm-leyline-utility Version: 1.0.0 The nm-leyline-utility skill is a decision-making framework for agent orchestration based on the research paper 'Utility-Guided Agent Orchestration for Efficient LLM Tool Use' (arXiv:2603.19896). It provides structured logic in files like SKILL.md and modules/action-selector.md to score candidate actions based on expected gain, cost, uncertainty, and redundancy. The skill is designed to optimize agent efficiency and cost-control, and it contains no evidence of malicious intent, data exfiltration, or unauthorized execution.