Wenxiao Zhang
HomePublicationsBlogProjectsSkills

Blog

Notes on LLM agents, retrieval systems, and engineering.

2026-06-14

Automatic Blog Posting with Codex

A short note on turning a blog idea into a published post through a Codex-assisted Markdown and API workflow.

  • blog-authoring
  • codex

2026-06-12

Surviving the Rebuttal: Why I Built Reviewvizskill

Drowning in a wall of reviewer comments during a rebuttal, I couldn't hold the thread or see what actually mattered, so I built a tool that turns the mess into a prioritized plan.

  • reviewviz
  • peer-review

2026-05-08

An Uncensored Coding LLM for the Pentest Lab

My supervisor wanted an LLM that would actually help students write exploit code in the penetration-testing unit he teaches. Mainstream models refuse, so we self-hosted one that doesn't.

  • pentestassist
  • llm

2026-02-18

STIndex: One Idea, One Week, One WWW Demo

Pascal pitched a sharp idea about spatiotemporal grounding for LLMs. With his guidance it came together in about a week (system, paper, and all) and landed in the WWW 2026 demo track.

  • stindex
  • spatiotemporal

2026-01-20

Building wenxiao.link

How this site works: Next.js + Django, publications auto-synced from Google Scholar with CORE ranks, served through a Cloudflare Tunnel.

  • meta
  • engineering