Archetype (A-16)

Archetype is my working library of GPT prompts and the place where I am progressively developing those prompts into GPT plugins.

I originally built this repository as a collection of standalone GPT prompt architectures: domain experts, creative practices, research systems, reasoning frameworks, personas, workflows, and other specialized archetypes. I am now updating selected prompts into more structured, inspectable, and reusable plugin packages while keeping the underlying prompt sources available in Markdown.

That means the repository currently represents both stages of the work:

• GPTs/ contains the broader GPT prompt library I have built over time.
• Prompts/ contains prompt sources I am actively organizing and evolving.
• Plugins/ contains packaged GPT plugins that I have already started producing from selected prompts.
• Construct/ contains my reusable Construct package and its underlying modules live in Prompts/Construct/.

I am treating this as an ongoing migration rather than a completed conversion. Not every GPT prompt has a plugin counterpart yet, and the prompt sources remain useful on their own.

Get Started · Repository Evolution · Packaged Systems · Construct · Construct Modules · Standalone Prompts · GPT Prompt Library · Repository Layout

I keep the repository intentionally inspectable. Most of the source material is plain Markdown, while the plugin distributions are versioned ZIP artifacts.

If you want to work with the prompt sources themselves:

• Browse my broader GPT prompt library in GPTs/.
• Use my Construct modules from Prompts/Construct/.
• Use Citation from Prompts/CT.md.
• Use Systematic Review from Prompts/SR.md.
• Use Visual Art from Prompts/VA.md.
• Browse additional standalone prompt specifications under Prompts/.

I use these Markdown files as the human-readable source layer. They can be supplied directly to an LLM or GPT environment wherever system prompts, custom instructions, or equivalent prompt content are supported.

I currently maintain four versioned packaged systems:

• Construct_v.0.9.18.zip
• Citation_v.0.1.1.zip
• Systematic_Review_v.0.1.7.zip
• Visual_Art_v.0.1.0.zip

I am using these packages as the next stage of the project: moving selected GPT prompts from standalone instructions into reusable GPT plugin form. Where I keep a corresponding Markdown source in the repository, I consider that source the easiest place to inspect the underlying prompt architecture.

I am evolving Archetype in roughly three layers:

1. GPT prompts — the original and still-growing collection of standalone prompt architectures in GPTs/.
2. Structured prompt sources — prompts I am refining into clearer source specifications under Prompts/.
3. GPT plugins — versioned packaged systems that I am building from selected prompt architectures under Plugins/ and Construct/.

I do not expect those layers to move in lockstep. Some prompts may remain useful as standalone GPT prompts, some may become structured source specifications first, and some may progress into packaged plugins.

My aim is to preserve the strengths of the prompt library while making the systems I continue developing easier to inspect, version, reuse, test, and package.

These are the systems I have packaged so far:

I expect this section to grow as I convert more of the prompt library into plugin packages.

Construct is my reusable core prompt architecture. I designed it as a set of general-purpose modules for reasoning, representation, knowledge work, judgment, modeling, learning, prioritization, validation, workflow design, and related tasks.

I keep the 20 source modules as standalone Markdown specifications under Prompts/Construct/, and I maintain a versioned packaged distribution at Construct_v.0.9.18.zip.

I use Construct as a shared conceptual foundation rather than as a single persona or task prompt. Each module defines a problem domain through distinctions, entities, relationships, operating models, evidence, uncertainty, governance, and workflow considerations.

• Reusable core: I use the modules for general reasoning and workflow primitives rather than a single domain or persona.
• Inspectable sources: I keep every core module readable as Markdown.
• Independent or combined use: I can use a module directly or package the system for environments that support it.
• Explicit boundaries: I define neighboring concepts separately instead of treating them as interchangeable.
• Evidence and governance: I treat provenance, uncertainty, validation, monitoring, and governance as first-class concerns where they matter.

I also keep several standalone prompt specifications at the top level of Prompts/:

• CT.md — Citation
• RS.md
• SR.md — Systematic Review
• ST.md
• TR.md
• VA.md — Visual Art

Some of these are already connected to packaged plugins, while others are still part of the source-side development process.

GPTs/ is the main library of GPT prompts I have created. It includes domain specialists, creative practices, structured workflows, perspectives, archetypes, and simulations of named people.

I am keeping this library available while I update selected prompts into newer source and plugin structures. I do not treat the older prompt format as obsolete simply because a plugin version may exist; the standalone prompts remain useful for direct inspection, experimentation, comparison, and reuse.

For named-person prompts, I use the name to indicate a simulation or perspective prompt rather than the actual person. I do not intend inclusion in the repository as an endorsement, and I do not treat a prompt's subject label as evidence of authority or factual accuracy.

archetype/
├── GPTs/                            # my broader GPT prompt library
├── Prompts/                         # source prompts I am organizing and evolving
│   ├── Construct/                   # 20 reusable Construct source modules
│   ├── CT.md                        # Citation source
│   ├── RS.md
│   ├── SR.md                        # Systematic Review source
│   ├── ST.md
│   ├── TR.md
│   └── VA.md                        # Visual Art source
├── Plugins/                         # packaged GPT plugins
│   ├── Citation_v.0.1.1.zip
│   ├── Systematic_Review_v.0.1.7.zip
│   └── Visual_Art_v.0.1.0.zip
├── Construct/
│   └── Construct_v.0.9.18.zip       # packaged reusable core
├── tests/
│   └── test_readme.py               # README local-link guard
├── .github/
│   ├── FUNDING.yml
│   └── workflows/
│       └── readme-check.yml         # runs the README link guard in CI
├── A_Avatar.png
├── LICENSE                          # Apache License 2.0
└── README.md

I am continuing to refine the prompt library and convert selected systems into GPT plugins. As I do that, I want the repository to preserve three things at once: the breadth of the original GPT prompts, the inspectability of their Markdown sources, and the versioned structure of the newer plugin packages.

The repository is therefore both an archive of my GPT prompt work and an active development space for the plugin systems that are growing out of it.