lcom-plugins

This repository is the working home for Learning.com context used by ChatGPT.

Its current primary deliverable is the lcom-business-context Skill, which provides curated business knowledge such as ontology definitions, glossary content, and business rules. The repository is intentionally structured so that additional Skills, Plugins, and/or MCP servers can be added later.

Repository purpose

Use this repository to:

• develop and maintain Learning.com business knowledge for AI use;
• create and test ontology, glossary, prompts, and supporting business context;
• maintain scripts and other tooling used to prepare or validate that knowledge;
• publish finished Skills, Plugins, and MCP components in locations intended for ChatGPT.

Repository structure

lcom-business-context/
├── README.md
├── workbench/
├── plugins/
├── mcp/
└── .agents/

workbench/

workbench/ is the main working area for humans.

Keep all development and maintenance material here, including:

• raw source material;
• ontology and business-knowledge drafts;
• prompts;
• Python maintenance scripts;
• test questions and expected results;
• notes and design documents;
• intermediate or generated files that are not part of a final ChatGPT component.

Example:

workbench/
├── raw/
├── drafts/
├── prompts/
├── scripts/
├── tests/
└── docs/

Files in workbench/ are not considered final ChatGPT deliverables.

plugins/

Reserved for final Plugin content intended for ChatGPT.

The current lcom-business-context Skill belongs under the appropriate Plugin structure here. Additional Skills or Plugins may be added later.

Only reviewed, ready-to-use content should be copied or promoted from workbench/ into plugins/.

mcp/

Reserved for MCP servers and related production-ready MCP components intended for ChatGPT use.

Development notes, experiments, prompts, and helper scripts should remain in workbench/ until they are ready to become part of an MCP implementation.

.agents/

Reserved for ChatGPT/OpenAI agent configuration and Plugin marketplace metadata.

Do not use this folder for general documentation, drafts, or maintenance files.

Working principle

The repository follows a simple separation:

workbench/
    = where humans create, investigate, draft, test, and maintain

plugins/
mcp/
.agents/
    = where finalized ChatGPT-facing components live

All new work should begin in workbench/.

Once content or code is reviewed and ready for use, move or copy the final version into the appropriate ChatGPT-facing destination:

• Skill or Plugin content → plugins/
• MCP server content → mcp/
• ChatGPT/OpenAI configuration or marketplace metadata → .agents/

Current scope

The repository currently focuses on the lcom-business-context Skill, including business ontology and curated business knowledge.

The structure is designed to support future expansion without requiring a new repository for every capability. Additional Skills, Plugins, MCP servers, and supporting AI components can be added as the project grows.

Maintenance guideline

When editing this repository:

1. Make changes in workbench/.
2. Review and test the result.
3. Promote only finalized content to plugins/, mcp/, or .agents/.
4. Keep ChatGPT-facing folders clean and free of drafts, raw material, and temporary files.