360 Secure SkillHub
360 安全技能中心

Skill Distiller (One-Liner)

该 Skill 在用户要求压缩、精简或降低某个 Skill 的上下文占用时触发;把目标 Skill 解析为章节,由 LLM 评估各章节重要性,在保留受保护模式(YAML name/description 等)的前提下删除低价值章节,最终给出压缩后的 Markdown、功能保留度评分、token 缩减统计以及删除/保留章节清单。自述为本系列的极简变体,功能保留度约 70%,评分由 LLM 估算。

leegitw
leegitw
热度 37.64
最近更新 2026-05-11 17:43:19
分类:效率依赖:内部调用大模型应用场景:摘要总结 / 工作流

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add neon-skill-distiller-oneliner
Metadata
name
Skill Distiller (One-Liner)
version
0.2.1
description
Skill compression reminder in 100 tokens — just trigger, action, result.
author
Live Neon <lee@liveneon.ai>
homepage
https://github.com/live-neon/skills/tree/main/skill-distiller/oneliner
repository
live-neon/skills
license
MIT
user-invocable
true
disable-model-invocation
true
emoji
🗜️
tags
["compression","skills","optimization","context-window","token-reduction","quick-reference","minimal","openclaw"]

Skill Distiller (One-Liner)

Minimal reference variant (~100 tokens, ~70% functionality, LLM-estimated). Full reference: ../SKILL.reference.md.

TRIGGER: User asks to compress, distill, or reduce a skill's context usage

ACTION: Parse skill into sections (TRIGGER/INSTRUCTION/EXAMPLE/etc), score importance via LLM, remove low-value sections while preserving protected patterns (YAML name/description, N-count tracking, task creation)

RESULT: Compressed skill markdown with functionality score (0-100%), token reduction stats, and list of removed/kept sections


Related

Variant Tokens Functionality
skill-distiller (main) ~400 ~90% (formula)
compressed ~975 ~90% (prose)
oneliner (this) ~100 ~70%

Full reference: SKILL.reference.md (~2,500 tokens, ~90%)

Functionality scores are LLM-estimated.