I'm the Technical Staff of Myself.
I build tools for AI agents — from knowledge visualization and modular skills to agent packaging and security infrastructure.
I work at the intersection of AI research, agent infrastructure, and hands-on tooling.
I map the AI landscape by turning 260+ industry reports into an interactive knowledge graph with 4400+ entities. I build infrastructure that makes agents portable and operational — from environment packaging to 17 modular skills covering search, scraping, memory, and security. And I dig into how leading agent frameworks work under the hood, contributing to OpenClaw and studying Claude Code's architecture.
Everything public here is MIT licensed.
Stack: Python · TypeScript · JavaScript · Shell · HTML/CSS 🦞
| Project | Description | Link |
|---|---|---|
| AI Knowledge Graph | Interactive visualization of the AI industry — 4400+ entities, 6200+ edges from 260+ reports | Repo · Demo |
| AgentBox | Portable AI agent configuration packaging and sharing tool | Repo |
| OpenClaw | Personal fork of OpenClaw — contributions and experiments | Repo |
| OpenCLI | Fork of OpenCLI — universal CLI hub for AI agents | Repo |
| Hermes Agent | Fork of Nous Research's Hermes Agent | Repo |
| Claude Code Source | Claude Code v2.1.88 source code — for technical research and study only | Repo |
Self-contained modules that extend AI agent capabilities. All built for OpenClaw.
| Skill | Description | Link |
|---|---|---|
| all-net-search-read | Search & extract from 7 platforms — WeChat (微信), Xiaohongshu (小红书), Twitter/X, YouTube, Reddit, Bilibili (B站), Weibo (微博) | Repo |
| web-access | Full internet access — 3-layer channel scheduling + browser CDP automation | Repo |
| book-hunter | Ebook search on Z-Library & Anna's Archive — title, author, ISBN with multi-layer fallback | Repo |
| wechat-reader | Read WeChat Official Account (微信公众号) articles as clean Markdown — no login required | Repo |
| Skill | Description | Link |
|---|---|---|
| x-twitter-scraper | X/Twitter public data — profile, timeline, full Note Tweets without login | Repo |
| weibo-scraper | Weibo (微博) public content scraper — visitor cookie mode, no login required | Repo |
| bilibili-video | Bilibili (B站) video download & transcription — no cookie, faster-whisper powered | Repo |
| xiaoyuzhou-podcast | Xiaoyuzhou (小宇宙) podcast fetcher — download, transcribe, summarize | Repo |
| Skill | Description | Link |
|---|---|---|
| ai-talent-graph | Academic AI scholar profiling — OpenAlex, arXiv, Semantic Scholar, ORCID integration | Repo |
| ai-talent-radar | Recruitment-oriented AI talent search — Semantic Scholar, GitHub, Zhihu (知乎), Weibo (微博) | Repo |
| ai-info-radar | AI news & resource discovery — 50+ sources with subscription push | Repo |
| agent-job-monitor | AI company Agent/LLM job monitoring — auto-collects from ByteDance, Tencent, Alibaba, Zhipu, Kimi, MiniMax | Repo |
| Skill | Description | Link |
|---|---|---|
| persistent-memory | 3-layer memory distillation — daily logs → structured facts → long-term memory | Repo |
| lobster-guard | Agent identity security — prevents impersonation & data leaks in group chats | Repo |
| agent-self-audit | Agent health check & diagnostics — memory, skills, cron, config | Repo |
| skill-security-guard | Skill code security scanner — 7 dimensions, A-F rating, prompt injection detection | Repo |
| skill-design-guide | Skill design, testing & QA methodology — distilled from 16+ production skills | Repo |