OpenAI terminal-based coding agent
Project homepage: https://github.com/openai/codex
Description
OpenAI Codex is a terminal-based coding agent.
Package information
- Ports path
- devel/codex
- Package architecture
- amd64
- Maintainer
- Kirill A. Korinsky <kirill@korins.ky>
- Categories
- devel, lang/rust
- Available flavors
- None listed
- Only for architectures
- amd64, aarch64
Ports list build and test dependencies as well as runtime dependencies. pkg_add selects those needed for installation.
Direct dependencies
Library
Runtime
Build
- devel/cmake/core
- devel/git
- devel/gn
- devel/ninja
- devel/protobuf
- databases/sqlite3
- devel/cargo-generate-vendor STEM->=1.1
- devel/llvm/22
- lang/rust
OpenBSD README
+----------------------------------------------------------------------- | Running codex on OpenBSD +----------------------------------------------------------------------- You may use OpenAI commercial services or misc/llama.cpp, for the last one you should add into ~/.codex/config.toml: model_provider = "local" model = "local" web_search = "disabled" [model_providers.local] name = "llama-server" base_url = "http://127.0.0.1:8080/v1" wire_api = "responses" codex uses a lot of memory, it starts with 256Mb per agent/worker thread, and you may need to bump your limits (`ulimit -d`) or raise datasize limit in used login class.. Also, codex may use a lot of open files, you may need to raise both the system-wide maxfiles and used login class limits.