Feature Suggestion: Enable Folder Path Indexing for Google Drive Integration
Context:
Currently, ChatGPT’s Google Drive integration indexes and retrieves files based on content and filenames, but completely ignores the folder structure. This means ChatGPT cannot see — or use — the parent folder name of a file, even when it’s semantically critical to disambiguating the content.
Why This Matters
Many users — especially those doing structured work like knowledge organization, data management, or experimental prototyping — use folder names as metadata.
Real-world example:
Imagine this structure:
people/
├── Andy/
│ └── height.txt → contains: "190cm"
├── Vicky/
│ └── height.txt → contains: "164cm"
└── Stan/
└── height.txt → contains: "177cm"
If a user asks, “What is Stan’s height?”, there is currently no way for ChatGPT to answer unless:
- The file content includes Stan’s name, or
- The file is renamed to something like
stan-height.txt
But this breaks the user’s clean design, where the folder is the label and the content is the value.
What’s Missing
ChatGPT does not currently index or expose the folder hierarchy, which makes it impossible to:
- Associate meaning from folder context
- Disambiguate identical filenames across folders
- Infer file ownership, category, or relationship when it’s embedded in directory structure
Proposal
Enable indexing of folder paths or at least immediate parent folder names for each file in Google Drive integrations.
This would allow ChatGPT to:
- Associate minimal content (like
"177cm") with the intended entity (Stan) - Support better file disambiguation
- Reflect how users naturally organize knowledge in folders
Benefits
- Significantly boosts inferencing quality
- Reduces friction for structured or minimal content use cases
- Aligns behavior with how users already structure data in Drive
Bottom line:
If you can’t see folders, you can’t think like the user.
Please consider adding support for Drive folder paths — it would unlock a much richer, more intuitive experience.