Codex Development Journal

Portfolio journal for documenting how projects were built with Codex: goals, prompts, outputs, problems, fixes, prompt revisions, final results, and lessons learned.

Features

• Records AI-assisted development sessions as structured JSON
• Tracks prompt revisions and lessons learned
• Renders polished Markdown case studies
• Helps demonstrate practical human-in-the-loop AI workflow
• Includes an example project journal

Architecture Diagram

flowchart LR
    A["Goal"] --> B["Prompt"]
    B --> C["Codex Output"]
    C --> D["Problem"]
    D --> E["Fix"]
    E --> F["Prompt Revision"]
    F --> G["Final Result"]
    G --> H["Lessons Learned"]
    H --> I["Markdown Case Study"]

Loading

Installation

python -m venv .venv
.venv\Scripts\activate
pip install -e .

Usage

codex-journal new ai-investigation-toolkit --goal "Build an investigator toolkit"
codex-journal entry ai-investigation-toolkit.json --prompt "Add IOC extraction" --output "Created regex extractor" --problem "Missed hashes" --fix "Added hash patterns" --lesson "Tests should include mixed IOCs"
codex-journal render ai-investigation-toolkit.json --out ai-investigation-toolkit.md

Roadmap

• Add Git commit linking
• Add before/after prompt diffing
• Add portfolio site export
• Add screenshots per journal entry

Known Issues

• The tool does not automatically import Codex transcripts.
• Entries are intentionally concise; long transcripts should be summarized before adding.

Contributing

See CONTRIBUTING.md.

License

MIT. See LICENSE.