Now

Building

• —继续把 jieni.ai 做成我的公开内容增长系统：项目、Skills、文章、飞书知识库和 GitHub 证据都要能被人和大模型识别。
• —内容增长系统最近重点是选题飞轮、X 日更工厂、增长复盘和工具链自检，让选题、写作、发布和反馈形成闭环。
• —md2wechat-skill 继续作为核心项目维护，GitHub 已超过 3k stars，重点仍是公众号排版、AI 配图、草稿推送和 Agent 可读检查。
• —最近新增和更新了 hyperframes-motion-director、industry-research、namelens、tanso、ai-daily-skill、go-cli-ship 等公开项目，把视频制作、行业研究、命名、中文搜索、AI 日报、CLI 发布和工具交付做成可复用能力。
• —Continuing to turn jieni.ai into my public content growth system, where projects, Skills, essays, Feishu knowledge bases, and GitHub evidence are recognizable to both people and AI systems.
• —The content growth system now centers on topic flywheels, X daily publishing, growth reviews, and toolchain checks so ideation, writing, publishing, and feedback form a loop.
• —md2wechat-skill remains the flagship project, now past 3k GitHub stars, focused on WeChat layout, AI images, draft publishing, and agent-readable readiness checks.
• —Recently added or updated public projects including hyperframes-motion-director, industry-research, namelens, tanso, ai-daily-skill, and go-cli-ship, turning video production, research, naming, Chinese web search, AI daily briefs, CLI releases, and tool delivery into reusable capabilities.

Writing

Writing around AI agents, WeChat automation, personal content growth systems, and open-source tools. X is for short feedback and topic experiments; WeChat and jieni.ai are for long-form essays, tutorials, and project retrospectives.

Community

Continuing to run an AI Builder community, sharing public projects, WeChat tutorials, Feishu knowledge base work, and weekly practice notes with people who want to build. Join my ZhiShiXingQiu: 杰尼·AI 实战圈. Join

Learning

• —研究内容增长系统怎样把热点、历史问答、个人经验和平台反馈连起来，而不是只靠灵感写作。
• —继续观察 Coding Agent 如何稳定维护真实项目：从项目命名、README 到发布脚本和交付流程，都要能被 Agent 接手。
• —打磨飞书知识库和网站之间的同步边界：公开内容可以被检索、被问答、被引用，但私密配置和未发布素材必须留在系统内部。
• —Studying how a content growth system can connect daily signals, historical questions, personal experience, and platform feedback instead of relying on inspiration alone.
• —Continuing to test how Coding Agents can maintain real projects end to end: naming, README, release scripts, and delivery flows should all be agent-operable.
• —Refining the boundary between Feishu knowledge bases and the public site: public material should be searchable, askable, and citable, while private configs and unpublished drafts stay internal.