Archify Tops Three GitHub Charts; Developer Tu Shaokun Looks Back on a Decade

Archify lets users tell AI about an event, a process or an unfinished idea and turn it into a clear diagram, Tu said. Users can also ask AI to revise missing or unclear parts. He gave examples such as planning a weekend schedule or photographing a hand-drawn sketch and asking Archify to organize it into a flow chart. The tool is not limited to software development; it can be used for product planning, business process design, event planning, content production and work handovers, he said. Some users embed Archify-generated dynamic charts into Feishu and DingTalk documents for communication or reporting, a use Tu said he had not expected.

The most frequent uses are recording daily life and sorting work processes, Tu said, because images make it easier to see stages, gaps and next steps than documents. Technical users also analyze unfamiliar repositories and projects by mapping complex structures. Beyond that, users have embedded HTML diagrams in collaborative documents, turned text or short essays into images, and converted photographed sketches into complete flow charts. Tu attributed Archify's popularity to real, tangible value: diagrams are more intuitive and easier to share than long text. He also placed dark- and light-theme examples and a demo video at the top of the GitHub README so visitors can immediately see what the tool does. In the future, he plans to fix bugs, polish user experience and carefully select community suggestions. For niche needs, he prefers guiding users to solve problems through AI dialogue rather than building heavy features such as web editing, and he may support importing common flow-chart formats if it brings value to a wider group.

Tu was born in Fuji Village, Xiangcheng, Zhoukou, Henan. He dropped out of school at 16 after a conflict with his father over mobile phone use, according to the interview. He went to Shanghai to learn car repair but left in less than a month because he could not lift tires; he then tried working as a waiter, on construction sites, as a building security guard and in a timber factory in Beijing and Ningbo, sometimes being dismissed, cheated or denied wages. He later returned to school. After the 2018 college entrance exam, he entered Chongqing College of Finance and Economics. His mother accompanied him more than 1,000 kilometers to Chongqing; they took a bus, an 18-hour hard-seat train and a high-speed train, arriving on the evening of the third day. He said his summer factory wages left little after tuition, so they chose the cheaper hard seat, which cost more than 120 yuan, while high-speed rail would have cost at least 600 to 700 yuan. He said a factory supervisor told him that people who enter factories are not all drifting, and a coworker who worked 12 to 14 hours a day while calling her children in her hometown left a deep impression. He also cited a classmate from a poor area of Guizhou who supported his family while studying and now owns a delivery station in Chongqing.

During his college years, Tu passed the college-to-university upgrade exam and entered Chongqing University of Posts and Telecommunications to study software engineering, ranking first in his batch under the guidance of teacher Fu Jian, according to the interview. He taught himself front-end development and accumulated internships. In April 2024, he interned at PatSnap in Suzhou, which he described as a happy period. But in autumn recruitment, he said, his resumes often received no response, and offers were withdrawn during background checks because of the label 'college-to-university.' At ByteDance's Feishu, he passed all interviews but the HR call ended after he wrote 'college-to-university' in the education field, and the offer never arrived. He sat by West Lake until sunset and his phone died. He later learned to tell HR upfront, and at an interview for a West Lake Future Gene internship, he burst out that 'a humble origin is not a shame' and received the offer. He also interned at SenseTime's large-model research institute. At Manus, he passed the technical interview but was blocked by HR over the same label; a second-round interviewer told him it was a pity and advised him to prove his ability through works or open-source projects. Tu said Manus's COO gave similar advice. He then joined the startup Zhiman Technology and now works at Xiaohongshu as an AI Native engineer.

Tu said open source is both self-serving and helpful to others: when a project genuinely helps people, it will receive sincere feedback. He said people in factories and vocational colleges also include positive, serious people, and no one should be dismissed solely by job or education. He advised people not to set limits on themselves and to try boldly regardless of major. 'Education, family background, or failing at something are not shameful,' he said. 'What is truly regrettable is giving up on ourselves.' QuantumBit reported that Archify had 3.8k forks and 58,800 stars as of publication, with a single-day peak of 5,000 new stars.