A way to exclude sensitive files issue still open for OpenAI Codex (github.com)

OpenAI Codex is facing ongoing challenges regarding the implementation of a feature that allows users to explicitly mark sensitive files or directories to be excluded from model access. The proposed solution includes creating a dual-layer configuration, consisting of a repository-specific .codexignore file and a global ignore file. This would prevent critical files such as environment variables and authentication keys from being inadvertently read or transmitted. Users highlighted the need for such a mechanism, particularly to safeguard sensitive data and to manage the exclusion of large or irrelevant files, ensuring that coding practices are both secure and efficient. This development is significant for the AI/ML community as it highlights the growing awareness of data privacy and security within machine learning environments. The ability to explicitly designate sensitive files supports a deterministic and shareable configuration that aligns with collaborative coding practices across teams. Although the discussion on this feature was previously closed in favor of a Rust-based implementation, its relevance persists as the demand for effective data management in AI applications continues to rise. The community is now eager to reignite this dialogue and converge on a functional design that meets these pressing needs.