Workflow & Core Tools · lijigang/ljg-skills

ljg-map

Name: ljg-map
Author: lijigang/ljg-skills

Analyzes user manuscripts through the epistemic lens of "Learning View", extracting concepts, criteria of distinction, and conditional connections to construct an accessible formalized knowledge map based on the manuscript's intrinsic relationships. Each node includes a plain-language definition, a concise formula, and concrete semantics, without imposing a preset hierarchy on the graph. USE WHEN the user invokes /ljg-map, or requests to map an article, note, or speech draft into a formalized map, cognitive terrain, or concept criteria and relationship diagram. NOT FOR standard summaries, paragraph-by-paragraph rewrites, geographic maps, or standalone formula proofs.

L1 Basic · Source registration onlyV1 Commit-PinnedDual-pass Audit

Author lijigangGitHub lijigang/ljg-skills ↗License MIT（仓库根 LICENSE）Commit 9e75497a95

Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

Core Mechanism: Leverages the concept / criterion / connection perspective of "Learning View" to discern the underlying knowledge structure of user manuscripts.

skills/ljg-map/SKILL.md

用《学习观》关于概念、判别模型和联结模型的理解，看到文稿背后的知识结构

What this does: Treats "Learning View" as an analytical lens (rather than output nodes) to identify concepts, criteria of distinction, and connections.

Workflow Routing: Read Workflows/Map.md when generating or revising a map; read Examples.md when calibrating expression scale.

skills/ljg-map/SKILL.md

从文稿生成地图，或修订已有地图

What this does: Routes execution to the main workflow document or example document based on request type.

The epistemic lens is merely a perspective: The alignment across concrete / conceptual / applied layers is an evaluation background, not fixed graph layers or steps.

skills/ljg-map/SKILL.md

具象、概念与应用的对应帮助检查知识是否接通，不构成固定的三个图层、三个栏目或必经步骤。

What this does: Mechanistically prevents the analytical framework from being mistakenly rendered as the fixed structure of the target object itself.

Nodes must originate from the current manuscript; applying rigid templates is prohibited.

skills/ljg-map/SKILL.md

**节点来自当前文稿。** 不把任何文章都画成「经验—抽象—概念—记忆—学习」

What this does: Enforces that the graph is determined by the material and that structure shifts with relationships.

Prioritize "Chinese Name + Short Formula" for formalization, without measuring by the sheer volume of symbols.

skills/ljg-map/SKILL.md

**形式化不以符号数量计。** 优先「中文名称 + 短公式」。

What this does: Suppresses excessive symbol stacking to ensure the map remains readable and comprehensible.

Semantic Discipline: Arrows carry semantic meaning; definition, classification, causality, and other relationship types must not be conflated.

skills/ljg-map/SKILL.md

**箭头有语义。** 定义、归类、计算、相关、因果、类比、推断和反馈不能混作一种关系

What this does: Enforces explicit distinction between relation types while preserving conditions and uncertainties.

Every node must be accompanied by a plain-language definition, a short formula, and concrete semantics (specified by the main workflow document).

skills/ljg-map/Workflows/Map.md

**白话定义**：它是什么，边界在哪里。

What this does: Mandates the minimal content set for node explanations so relationships can be understood and verified.

Delivery Convention: Deliver Markdown + Mermaid + readable formulas within the conversation by default; do not write to notes or publish by default.

skills/ljg-map/SKILL.md

默认直接在对话中交付 Markdown、Mermaid 和可读公式。

What this does: Clarifies delivery boundaries, restricting file writing or publishing to non-default actions upon explicit user request.

Read-back Standard: Read back the graph, node explanations, and original text before delivery; do not treat valid syntax as visual acceptance.

skills/ljg-map/Workflows/Map.md

能解析的图不等于正确或易读的图。

What this does: Mandates checking main arguments, branches, and edge direction conditions, avoiding relying solely on passing Mermaid syntax validation.

2Core Capabilities

01Generate formalized knowledge maps from manuscripts/notes/lecture drafts, or revise existing maps

02Extract concept referents, criteria of distinction, and conditional connections

03Provide a plain-language definition, short formula, and concrete semantics for each node

04Select an appropriate graph topology based on content (set intersection / conservation / conditional branching / feedback loop / relational network)

05Relationship semantic validation: Distinguish definitions from regularities, necessity from contingency, and avoid escalating correlation to causation

06Revise existing maps: Reduce symbol clutter and main diagram crowding using a single cohesive example throughout

07Restrained delivery boundaries: Output exclusively within dialogue by default; refrain from unauthorized note-writing, overwriting sources, or publishing

4Recommended Governance & Evidence

Suggested Controls for Safe Operation

For Agent Runtime

Can be auto-loaded

For Administrators

Evidence & Triggers (File Paths & Excerpts)

SKILL.md: Pure prompt instructions and rule text; no script execution or external impact

Audit Notes: Pure prompt + reference document workflow: All four files in the directory are markdown/yaml text with no scripts, no network endpoints, and no credential access, rated as green. It instructs reading user manuscripts and optional environment source tools for calibration, which constitutes host read/tool actions and does not change the rating category per review criteria.

5Second-Pass Verification

[ok] 四文件角色与 internal_assets — SKILL.md（入口）+ Workflows/Map.md（主流程）+ Examples.md（样例）+ agents/openai.yaml（界面元数据）逐一可定位，路径与 §1 清单一致。
[ok] 反模板化纪律一致 — SKILL.md「知识观是洞察视角，图形由材料决定」与 Workflows/Map.md「不是排版要求」相互印证。
[ok] 节点解释最小集 — Workflows/Map.md「每个节点的可用定义」给出白话定义/短公式/具体语义/必要边界四项，与 SKILL.md 成品约定一致。
[ok] 无脚本/网络/凭证 — 四文件均为 markdown/yaml 文本，风险扫描六类零命中，scripts_executed/network_calls/credential_reads 据实留空。
[discrepancy] 校准取数未在扫描中体现 — 风险扫描六类零命中，但 Workflows/Map.md 允许用「当前环境可用的来源工具」做事实校准；该调用非 skill 自带依赖、工具与出网范围由宿主决定，故 external_deps 留空并在此记录。
[ok] 读回标准已纳入 — Workflows/Map.md「读回标准」要求交付前读回图/节点解释/原文，并声明语法通过≠视觉验收，已写入 how_it_works 第 9 条。

6Verdict & Scenarios

The principle of separating the perspective from the target object is repeatedly emphasized, mechanistically preventing the analytical framework from being misdrawn as the object's fixed structure.

Semantic rigor: Distinguishes classification from prediction, enforces semantic arrows, marks gaps where necessary, and avoids upgrading correlation to causation.

Restrained delivery with clear boundaries: Defaults to conversation output containing Markdown + Mermaid, avoiding unauthorized file writing or publishing.

Includes self-contained example documents demonstrating structure varying with relationships using two fictional passages, accompanied by counterexamples of expression scale.

Best for: Ideal for researchers, instructors, and writers who need to visualize the argumentative structure of articles/notes/lecture drafts, especially users who prioritize conceptual distinctions and relational conditions over conventional summaries.

Not for: Not suitable for: Users seeking generic summaries or paragraph-by-paragraph paraphrasing; scenarios requiring geographic maps or isolated formula proofs; host environments unable to render Mermaid or requiring direct output of images/webpages; and users expecting the skill to automatically write note files or publish artifacts.

Install Agent-Ready

① Verified Mirror 2026-10-10

Option A · Direct DownloadDownload ljg-map.tar.gz

sha256: 00bf29b976102eb6…

Option B · Copy Guide for Agent

install.json

Agents automatically download and verify mirrors from this JSON guide.

② Upstream GitHub

GitHub Repo ↗

Pinned to commit 9e75497a95.