During November 2024, Afierka enhanced the HabanaAI/vllm-hpu-extension repository by delivering a focused documentation update supporting FP8 inference workflows on HPU. Leveraging Markdown and model calibration expertise, Afierka clarified the FP8 inference process with vLLM and Intel Neural Compressor, refining command-line argument explanations and adding targeted warnings and tips to reduce user misconfigurations. The work improved onboarding speed and setup reliability for new users and contributors, aligning documentation closely with actual workflow requirements. While no bugs were addressed, the update strengthened maintainability and laid groundwork for future automation and CI integration, demonstrating depth in documentation and workflow evaluation.