Guides
Start typing to search…

How to make your project searchable with the Claude Code plugin

This article covers how to upload your project files so you can search them, and how to keep them up to date. For background on how it all works, see How the Claude Code Plugin Works.

Upload your project for the first time

  1. Open your project in Claude Code.
  2. Run /create-agent.
  3. The plugin scans your folder, uploads all eligible files to a new CustomGPT.ai agent, and saves the connection to .customgpt-meta.json at your project root.
  4. Run /check-status and wait until index_status shows ok - that means your files are ready to search.
  5. Once ready, see How to Search and Manage Your Project to start asking questions.

Upload taking a long time? Large projects can take several minutes. Keep running /check-status to check progress. If index_status never changes to ok, check your API quota in the CustomGPT.ai dashboard.

What gets uploaded? Code files, config files, documents (PDF, DOCX, XLSX, CSV, and more), and plain text. Excluded automatically: .git/, node_modules/, __pycache__/, dist/, build/, .next/, .venv/, vendor/, and .env files. See Supported file types for the full list.

Keep your files up to date after changes

Use the right command for the scale of change:

What changedCommand to use
One file/update-agent path/to/file.py
Several files or a folder/add-files src/
Large refactor or files feel out of date/rebuild-agent

/rebuild-agent removes everything and re-uploads your whole project. Use it after major changes or if search results feel stale.

📘

Freshness warning: If you search after files have changed, the plugin warns you: "X file(s) have changed since the last upload. Results may not reflect the latest changes. Run /rebuild-agent to re-sync."

Add specific files without re-uploading everything

Run /add-files followed by a file path or folder:

/add-files docs/architecture.md
/add-files src/

If no agent exists yet, /add-files creates one automatically.

Add images

Images (.jpg, .jpeg, .png, .webp) can be added via /add-files and /rebuild-agent, but not /create-agent. To include images, run /add-files after the initial setup is done.

When image files are detected, the plugin asks:

Image files detected (3 images). Enable AI Vision processing? (yes/no)
Compress images before vision processing? (yes/no)

AI Vision extracts content from diagrams, screenshots, and scanned documents.

Supported file types

Code: .js .ts .tsx .jsx .py .go .rb .java .cs .cpp .rs .swift .kt .php .sh .html .css .scss .vue .svelte .sql .graphql .proto and more

Config / Data: .json .yaml .toml .xml .ini .cfg .env.example

Docs: .md .rst .txt .pdf .docx .pptx .xlsx .csv

Images: .jpg .jpeg .png .webp

Related articles

Updated about 2 hours ago