Project Forge gives you a deliberate, structured way to open a ChatGPT Project. Instead of hoping the model remembers your last conversation, you prepare an explicit opening package — files the target project reads at startup — so every session begins from a known, verified basis.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/XxYouDeaDPunKxX/GPT-PF-Chat-GPT-Project-Forge/llms.txt
Use this file to discover all available pages before exploring further.
Introduction
Understand what Project Forge is, what problem it solves, and when to use it.
Quickstart
Set up your first opening package and launch a target project in minutes.
Core Concepts
Learn the information model, authority hierarchy, and readiness states that govern the system.
Operator Guide
Practical operating rhythm — session setup, hygiene, failure modes, and smoke testing.
How it works
Project Forge separates the system’s stable rules from the per-project state it prepares. The canonical core never holds live case data — that lives in external artifacts the target project reads at startup.Load the canonical files
Provide
AI_START.md and the four canonical files (00_SCOPE.md, 01_RULES.md, 03_ARTIFACTS.md, 02_PROTOCOL.md) to your AI session.Close the target
Define the objective, scope in, and scope out for the project you want to open. This is the foundation of your opening package.
Compile the artifacts
Generate the minimum opening package: an
ENTRY_POINT.md, a TARGET_PROJECT_BRIEF_ARTIFACT, and an INITIAL_SSOT_ARTIFACT. Add a handoff or transfer artifact only if your case genuinely needs them.Explore the documentation
Using Project Forge
Manual, guided, agent-based, and web compiler workflows.
Canonical Reference
The four governing files that define scope, rules, artifacts, and protocol.
Artifact Templates
Schemas and field definitions for every artifact class.
Project Forge is licensed under CC BY-SA 4.0. It was developed with AI assistance under human direction.