haorantang97

@haorantang97 source repo

17 skills
0 followers
20 hours ago last updated

▌
Tephra · haorantang97 bundle
把一张图重画成黑底像素画，两种正典画法按题材二选一：**纯粒子**（会动的短笔触沿图像自身的结构方向流动，适合浪／烟／光柱这类本来就在流动的东西），或**网点实体**（把明暗翻译成逐帧静止的有序点阵，适合人像／雕塑／动物这类要读出轮廓和体积的东西，并且只有它能把形体梳成硬边碎裂条）。动效按画面的物理规律分区规划；可选地让一部分形体溃散成飞灰，同时锚定另一部分保持实体与原色。输出静态 PNG 与无缝循环 GIF/MP4。触发：/tephra、「粒子流动效果」「把这张图重绘成笔触」「黑底像素画」「网点碎裂」。算法与代码全部原创，输出不基于任何模型，也不复制任何具体作品。

0
installs
▌
Antibes Holiday · haorantang97 bundle
Create original relaxed black-pen graphics from concepts or visual references, including sparse illustrations, narrative scenes, animals, objects, abstract relationships, icons, and logo marks. Use when the user wants quick hand-drawn line energy, causal story staging, non-equilibrium proportion, shorthand forms, open contours, structural line drift, misregistered junctions, selective retracing, physical pen texture, or a transferable illustration system that must not collapse into clean vector minimalism or polished period-sketch drawing.

0
installs
▌
Lab Life Reviewer · haorantang97 bundle
Use when a user wants a long-running personal life review, autobiographical interview, life-event reconstruction from related materials, detailed handoff, or approved transfer into long-term personal context. / 用于长期人生回顾、个人经历访谈、结合相关材料还原事件、生成详细交接包，以及经批准后写入长期个人上下文。

0
installs
▌
Lab Knowledge Intake · haorantang97 bundle
Use whenever the user says 录入知识库、导入知识库、保存到知识库、沉淀到知识库, asks to add/update/merge/move/delete knowledge, or provides a link, attachment, text, or conversation to put into the user's schema-governed Markdown/Obsidian knowledge base through Agent Knowledge.

0
installs
▌
Lab Context Distillation Wx · haorantang97 bundle
Use when a user wants to collect, snapshot, lawfully decrypt, normalize, redact, map, merge, audit, adjudicate, or distill local macOS/Windows WeChat 4.x conversations into evidence-bounded personal context. Also trigger for 微信聊天蒸馏、微信数据库解析后的蒸馏、微信聊天知识库、全量 Map、跨事件归并、冲突补漏、蒸馏验收 or resumed WeChat distillation runs. Do not use for interview-led life review or general knowledge intake.

0
installs
▌
Lab Knowledge Retrospective · haorantang97 bundle
在一项实质性任务、事故、访谈或长对话结束、暂停或被判定失败后，做证据化复盘，并把通过门槛的候选结论交给提案审批流程。既支持短任务的结论复盘，也支持对失败项目、长对话、多个关联任务以及 “深度复盘”“逐次问答复盘”“为什么偏航/漏答”的逐轮法证审计；后者必须读取原始轮次、核验完成声明、关联用户纠正并通过覆盖门，不能用最终状态摘要代替过程审计。用于“复盘一下这轮”“总结这轮” “这次为什么失败”“学到了什么”“刚才那个坑记一下”“有哪些经验值得以后复用”等请求。 Use whenever a finished or explicitly paused task, incident, interview, or long conversation should be audited and distilled into evidence-bounded, reusable conclusions. 不得用于仍在推进且当前目标是完成修复的故障排查、单纯保存进度或普通内容摘要。本 skill 不直接写 Vault 或派生索引；需要入库时先取得 agent-knowledge 的当前 intake 契约，并严格走精确提案与用户审批流程。

0
installs
▌
Open Loops · haorantang97 bundle
当用户希望找回长对话中没有被看见、确认或明确取舍的信息点时使用。包括四类情况：用户提了但 AI 没答、AI 提了但用户没接住、AI 替用户做了选择，以及 AI 本该提问却自行决定。用户说「对账」「建清单」「哪些没闭环」「open loops」，或抱怨对话跑偏、旧问题消失、AI 替自己做主时触发。长项目进行到一半才开始审计，也属于正常入口。用户明确启用后，在该项目或任务中持续逐轮运行，直到用户明确暂停或停止，不要求每轮重新点名。不要用于单轮编号问答、普通任务进度追踪，或用户已经明确看见但暂缓执行的事项。仅检查、调试、修改或发布 Open Loops 本身的轮次属于维护通道：可以使用本 skill 排障，但不得写进业务项目的原文存档或未闭环清单。

0
installs
▌
Skill Skill · haorantang97 bundle
Create, revise, validate, package, or publish AI skills and rule files. Use as the main entry for the user's skill-authoring workflow, including creating a new SKILL.md, improving an existing rule, and preparing a skill repository. Choose only the requested stage; ordinary product code or content work is outside this skill.

0
installs
▌
Super Astra · haorantang97 bundle
Initialize Super Astra and select useful Astra guidance for ordinary task assignments, implementation, research, writing, questions, and follow-ups. Use when installing or first invoking the skill, when the host router supplies its catalog, or when a user asks to adjust agent behavior. Select modules by task intent before working; no complaint or keyword is required. Simple requests can use no modules. Do not turn ordinary tasks into configuration audits.

0
installs
▌
Reading Plan Mentor · haorantang97 bundle
把一份书单变成一个长期的、有导师陪读的每日阅读计划，并自动定时投递（默认邮件，可切换）。当用户提到"读书计划""共读/陪读""想系统读完这几本书""给一份书单帮我安排""每天给我导读" "长期学习/阅读计划""把读书做成每日推送""读书打卡"等，或上传/列出多本书并问怎么安排时，都应使用本 skill——即使他们没有明说"skill"或"每日邮件"。本 skill 覆盖：主题识别（含诚实退出）、按书性质分类（论著/随笔/小说等线性叙事）、按时长的难度加权容量与节奏控制、跨文本钩子、从日常对话里抓取并压缩用户的思考、进度看板与阶段复盘。 Use whenever a user wants to design or run a multi-day/multi-week reading plan, a daily reading guide, a read-along companion, or wants a booklist turned into a paced schedule with recurring deliveries — even if they don't say "skill" or "daily email".

0
installs
▌
Write Readme · haorantang97
Write or renovate a repository README that explains its purpose, helps readers choose and use its tools, and matches its visual identity. Use for a new repository introduction or an existing landing-page improvement; preserve a narrow edit's scope.

0
installs
▌
Translate Readme · haorantang97
Use this skill when the user wants a bilingual or multilingual README for their AI skill or rule repository. This includes translating README.md into a second language as a separate file (README.zh-CN.md, README.ja.md), adding the language switcher line to the top of both files, and keeping the two versions in sync after edits. Trigger when the user says 'add a Chinese README', 'make my repo bilingual', 'translate the README', 'add 中文 docs', 'I want English and Chinese versions', or 'add a language toggle'. Also trigger when the repo targets a bilingual audience and the README exists in only one language. Do NOT trigger for translating the rule file or SKILL.md content itself (rule files stay in one language so agent behavior is predictable), or for writing the original README from scratch (that is write-readme), or when no README exists yet.

0
installs
▌
Publish To Github · haorantang97
Use this skill when the user is ready to push their local AI skill or rule repository to GitHub for the first time. This includes running git init and the initial commit, creating the GitHub repository via the gh CLI or manual browser steps, setting the repository description and Topics tags, and confirming the push succeeded. Trigger when the user says 'upload to GitHub', 'push this to GitHub', 'create my repo on GitHub', 'how do I publish this', 'make this live', or 'I have all the files ready, now what'. Also trigger when the user has a complete local folder with all files and just needs the final publishing step. Also covers environments with no usable git at all — sandboxed agents, browser-only sessions — via the GitHub API with a fine-grained PAT, browser drag-and-drop upload, agent-driven web editor commits, and a repair mode for repos that already exist. Do NOT trigger if any files still need to be written or edited before publishing — all content must be finalized first. Do NOT trigger for submitting

0
installs
▌
Polish Rule Content · haorantang97
Revise an existing AI skill or rule for clearer discovery, useful instructions, appropriate scope, and reliable execution. Use for a requested content improvement or a demonstrated behavioral problem; a review-only request produces findings without rewriting.

0
installs
▌
Scaffold Repo Files · haorantang97
Use this skill when the user needs to create the supporting files for a GitHub repository beyond the rule file itself and the README. This includes choosing and writing the LICENSE file, writing CONTRIBUTING.md with submission format and quality gates, creating .github/PULL_REQUEST_TEMPLATE.md, setting up the correct directory structure, and creating optional files like CHANGELOG.md and SECURITY.md. Trigger when the user says 'set up my repo files', 'what other files do I need', 'create a CONTRIBUTING file', 'scaffold the repo structure', 'set up the directory layout', or 'I have the skill and README, what's next'. Also trigger when the user is about to publish and asks what supporting files a proper repo should have. Do NOT trigger for writing the README itself (that is write-readme), for creating banner images (that is create-visual-assets), or for the git and GitHub operations (that is publish-to-github).

0
installs
▌
Create Visual Assets · haorantang97
Create or revise a repository's hero, theme variants, social preview and badge row. Use for repository visual assets or a README renovation; apply the supplied reference and product identity. Use an appropriate image tool for raster illustration.

0
installs
▌
Submit To Directories · haorantang97
Use this skill when the user wants to submit their published AI skill or rule repository to community directories and awesome-lists to increase visibility. This includes identifying which directories accept the repo's format, formatting the entry correctly per each directory's CONTRIBUTING.md, and walking through the fork → add entry → open PR process. Trigger when the user says 'submit to awesome list', 'how do I get my skill listed', 'add to the community directory', 'submit a PR to awesome-cursorrules', 'how do people find my skill', or 'promote my repo'. Also trigger right after the repo has been published and the user asks what to do next. Do NOT trigger before the repo is live on GitHub — the repo URL must exist before submitting anywhere. Do NOT trigger for the Claude Plugin Marketplace submission process, which requires a separate .claude-plugin/marketplace.json configuration.

0
installs

17 published skills