# Adaptive Prompt Architect — Agent Build Stack

This package contains build materials for creating the **Adaptive Prompt Architect** agent across:

- ChatGPT GPT Editor
- Claude Projects / custom instructions
- Gemini Gems / custom instructions
- Gemini inside NotebookLM as a role/style header

## Files

| File | Purpose |
|---|---|
| `01_APA_Master_System_Header.md` | Canonical full system header / behavior specification |
| `02_GPT_Editor_Build.md` | GPT Editor-ready build instructions |
| `03_Claude_Project_Instructions.md` | Claude-ready instruction packet |
| `04_Gemini_Gem_Instructions.md` | Gemini/Gem-ready instruction packet |
| `05_NotebookLM_Role_Header.md` | NotebookLM role/style header version |
| `06_Output_Templates.md` | Prompt packet templates by task type |
| `07_Calibration_Tests.md` | Test prompts and expected behavior |
| `08_Registry_Record.json` | Machine-readable agent registry record |
| `09_Quickstart_Operator_Guide.md` | How to use the agent in practice |

## Core Identity

**Adaptive Prompt Architect** converts loose, messy, long-context, serious-project operator input into copy-ready, production-grade prompts for advanced reasoning models, multimodal agents, coding agents, research systems, and automation tools.

It sits between everyday prompt generation and full multi-agent orchestration.