Codeium provides AI-powered code completions and can understand your project’s codebase to offer contextual suggestions. When Codeium does not recognize your custom code libraries, the LISBOA77 suggestions become generic and miss the project-specific functions, classes, and modules that you use daily.
Here is why this happens and how to improve recognition.
Why Custom Libraries Go Unrecognized
Codeium’s workspace indexing may not have reached your custom library files. If the library is located outside the active workspace directory, or if indexing was incomplete, those files will not be part of the context.
Custom libraries stored in dependency folders like node_modules or virtual environments may be excluded from indexing by default. Editors commonly ignore these directories to reduce scan time, and the AI tool inherits this exclusion.
If your library is private and not published to any public registry, Codeium’s general knowledge base will not include it. The tool can only suggest based on what it has seen in its training data or what is indexed locally.
Outdated workspace indexes do not reflect recent changes to your custom library. If you added new functions or classes since the last index, those additions will not appear in suggestions.
Quick Fixes to Try
Open the folder containing your custom library in the same workspace as your main project. This ensures Codeium scans those files during indexing.
Trigger a re-index of your workspace. Look for a command in Codeium’s settings or command palette to refresh the workspace index. This forces the tool to re-scan all accessible files.
Check your editor’s file exclusion settings. Make sure the directory containing your custom library is not listed in the excluded files or folders configuration.
Import your custom library explicitly in the files where you use it. Clear import statements give Codeium stronger signals about which modules are relevant to the current file.
Advanced Solutions
If the library is in a separate repository, consider adding it as a symlink or a local dependency that your project references directly. This brings the library files into the workspace structure.
Add JSDoc, docstrings, or type annotations to your custom library code. Detailed documentation within the source files gives Codeium more context to generate accurate suggestions.
Check whether Codeium supports workspace-level configuration files where you can specify additional directories to include in the index.
If you use a monorepo structure, make sure Codeium is configured to scan the full repository rather than just the currently active subfolder.
Protecting Your Code
If your custom library contains proprietary code, review Codeium’s privacy and data handling policies before enabling full workspace indexing. Understand whether your code is sent to external servers for processing or handled locally.
Consider using Codeium’s enterprise or self-hosted options if data privacy is a priority for your organization.
In Summary
Codeium not recognizing custom code libraries is typically an indexing scope or workspace configuration issue. Including the library in your workspace, triggering a re-index, and adding clear documentation to library files will significantly improve recognition and suggestion quality.
