What Is GLOSSARY.md in Cangjie‑skill? Understanding the Central Terminology Dictionary

GLOSSARY.md is the global terminology dictionary for the cangjie‑skill repository, collecting all terms and definitions from across skill modules into a single, shared reference file generated during the Zettelkasten workflow stage.

In the cangjie‑skill framework, terminology consistency and discoverability are critical for managing multiple skill modules. The GLOSSARY.md file addresses this by serving as a central lookup point for every concept, term, and definition extracted from the source knowledge base. Unlike individual skill definitions that describe behavioral logic, this file focuses purely on shared vocabulary.

What Does GLOSSARY.md Contain?

The glossary extractor generates GLOSSARY.md by scanning source materials for candidate terms. According to methodology/02-stage1-parallel-extract.md, the glossary-extractor operates as one of several parallel extractors that identify key concepts during the initial processing stage.

Each entry in GLOSSARY.md includes:

• Term name — the concept or keyword
• Concise definition — explanation suitable for cross‑reference
• Context links — optional references to where the term appears

This structure ensures that when multiple skills reference the same concept, they use identical definitions, reducing ambiguity and maintenance overhead.

Where GLOSSARY.md Lives in the Workflow

The file's location changes during the build pipeline. Per methodology/05-stage3-zettelkasten.md, the workflow moves the temporary extraction output to its final position:

# After parallel extraction completes

mv candidates/glossary.md books/example/GLOSSARY.md

This placement at the repository root makes it accessible to all skill modules regardless of their folder depth.

Final delivery links the glossary from INDEX.md. As documented in methodology/07-stage5-deliver.md, the generated index includes:

# INDEX.md (auto-generated)

- [Skill Map](SKILL.md)
- [Global Glossary](GLOSSARY.md)

Readers can thus navigate directly to terminology definitions from the main entry point.

How GLOSSARY.md Differs from Skill Definitions

Aspect	GLOSSARY.md	*/SKILL.md
Purpose	Global term dictionary — defines what things mean	Individual skill description — defines what skills do
Granularity	One entry per term, shared across all modules	One file per self-contained skill module
Location	Root of output repository	Inside each skill's dedicated folder
Generated by	glossary-extractor (parallel extraction stage)	Framework, principle, case, and other specialized extractors
Linked from	INDEX.md as global reference	INDEX.md skill map and internal skill TOC

A skill definition in SKILL.md focuses on behavioral logic: trigger sentences, input/output formats, and execution flow. It may reference glossary terms but does not define them. For example:

<!-- Inside books/example/skill/SKILL.md -->

### Example Skill: 解释文字

本技能接受用户的文字描述，并返回对应的 **概念解释**。  
使用的术语请参见全局词典：[`GLOSSARY.md`](../GLOSSARY.md) → *"概念解释"*。

This separation of concerns — definitions in one place, behavior in another — keeps the repository maintainable as it scales.

Key Implementation Files

Understanding GLOSSARY.md requires familiarity with these source files in kangarooking/cangjie-skill:

• extractors/glossary-extractor.md — implements term identification and extraction logic
• methodology/02-stage1-parallel-extract.md — documents the parallel extraction stage where the glossary extractor runs
• methodology/05-stage3-zettelkasten.md — describes moving candidates/glossary.md to the final location
• methodology/07-stage5-deliver.md — specifies linking the glossary from INDEX.md
• GLOSSARY.md — the generated output file itself
• SKILL.md — template showing how skills reference glossary entries

Summary

• GLOSSARY.md is the central terminology dictionary for the entire cangjie‑skill repository, generated by the glossary-extractor
• Location workflow: extracted to candidates/glossary.md, then moved to repository root during Zettelkasten stage
• Cross-linking: INDEX.md provides global navigation to the glossary for reader discoverability
• Separation of concerns: glossary defines what terms mean; individual SKILL.md files define what skills do
• Consistency benefit: shared vocabulary prevents drift when multiple skills reference identical concepts

Frequently Asked Questions

How is GLOSSARY.md generated in cangjie-skill?

The glossary-extractor processes source materials during the parallel extraction stage (methodology/02-stage1-parallel-extract.md), identifying candidate terms and writing them to candidates/glossary.md. The Zettelkasten stage then moves this file to the repository root as GLOSSARY.md.

Can skills reference terms from GLOSSARY.md?

Yes. Individual SKILL.md files use relative links like [GLOSSARY.md](../GLOSSARY.md) to point readers to term definitions. This keeps skill descriptions focused on behavior while maintaining terminological precision.

What happens if the same term appears in multiple skills?

All instances point to the single definition in GLOSSARY.md. The extractor consolidates duplicate candidates, ensuring one authoritative entry per concept regardless of how many skills use it.

Is GLOSSARY.md manually edited or fully automated?

The file is generated automatically by the glossary extractor, though the methodology allows for manual curation of candidates/glossary.md before the move operation if refinement is needed.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}