codex-workflows and codex-spec
Tool A provides production-ready development workflows powered by AI subagents for Codex CLI, while Tool B offers automated, spec-driven development workflows specifically for OpenAI Codex, making them complementary tools that could be used together to enhance different aspects of a Codex development pipeline.
About codex-workflows
shinpr/codex-workflows
Production-ready development workflows for Codex CLI, powered by specialized AI subagents.
About codex-spec
shenli/codex-spec
Automated workflows for OpenAI Codex. Features spec-driven development for new features: Product Requirement → Design → Development Plan → Implementation.
Maintains shared project context (product requirements, tech stack, codebase structure) as evolving markdown sources, enabling Codex to generate aligned specifications, requirements, and task breakdowns with automatic dependency tracking. Executes tasks with full workspace context through a sandbox layer that preserves implementation history in `tasks.json`, while `--auto` mode uses `git diff` to incrementally refresh context as code changes. Built as a Node.js CLI that orchestrates OpenAI API calls to drive the full workflow from intent through phased task execution and status reporting.
Scores updated daily from GitHub, PyPI, and npm data. How scores work