Machine-readable resume — fetched and edited by AI agents (Claude, Cursor, ChatGPT) through the cv-pro CLI or MCP. Also served as raw JSON at /megan.json.
Copy {
"username": "megan" ,
"header": {
"name": "" ,
"tagline": "AI 原生流程产品经理（零售运营方向）"
} ,
"personalInfo": {
"email": "" ,
"location": "广州"
} ,
"experience": [ ] ,
"education": [
{
"school": "" ,
"major": "数据科学（计算语言学 / NLP 方向）" ,
"degree": "硕士" ,
"startDate": "2021-08" ,
"endDate": "2023-07"
}
] ,
"projectsRecent": [ ] ,
"projectsDetailed": [
{
"title": "Text2SQL / Data Agent Prototype" ,
"type": "可验证的数据 Agent 工作流" ,
"startDate": "2025" ,
"endDate": "2026" ,
"url": "" ,
"award": "" ,
"bullets": [
"基于 LangGraph StateGraph 设计自然语言到 SQL 的 Agent 工作流，覆盖意图识别、澄清、Schema 检索、生成、校验、人审、执行、重试、答案合成和审计日志。" ,
"设计“模型生成 + 确定性校验 + 人类审批”边界：模糊意图触发澄清，高风险操作进入人工确认，执行失败进入重试和错误注入修正。" ,
"将 Agent 可靠性拆成状态路由、工具输入、失败路径、用户确认和日志沉淀，可迁移至零售运营中的库存查询、异常解释、规则判断和流程辅助场景。"
] ,
"tags": [
"LangGraph" ,
"StateGraph" ,
"Schema Retrieval" ,
"Human Approval" ,
"Audit Log"
]
} ,
{
"title": "Skill Lens / Agent 工作流评估" ,
"type": "失败路径与边界审计方法" ,
"startDate": "2026" ,
"endDate": "2026" ,
"url": "" ,
"award": "" ,
"bullets": [
"构建面向 Agent / Skill 的 5 维度评估框架：架构设计、决策节点、失败路径、提示词/接口边界、责任边界，用于判断 AI 是否真正进入任务链路。" ,
"评估维度从单轮答案质量扩展到任务理解、工具调用、结果校验、失败恢复、用户确认与产物保存，适配真实运营流程中的异常分支、权限边界和不信任场景。"
] ,
"tags": [
"AI 产品评估" ,
"Badcase 复盘" ,
"责任边界"
]
} ,
{
"title": "科学营养顾问 Skill" ,
"type": "AI Skill 产品化与状态管理" ,
"startDate": "2026" ,
"endDate": "2026" ,
"url": "" ,
"award": "讯飞 ClawHub 二等奖、专区 Top 8、OpenClaw 400+ 下载" ,
"bullets": [
"覆盖 12+ 健康管理场景，设计用户档案、本地日志、今日剩余看板和场景触发词体系，将一次性问答扩展为可持续使用的 AI 工作流。" ,
"获讯飞 ClawHub 二等奖、专区 Top 8，OpenClaw 400+ 下载，证明从场景拆解、交互脚本、状态管理到平台发布的 AI 产品化能力。"
] ,
"tags": [
"讯飞 ClawHub" ,
"OpenClaw" ,
"用户状态" ,
"本地日志" ,
"场景触发词"
]
}
] ,
"skills": [
{
"name": "产品能力" ,
"items": [
"AI 场景判断" ,
"流程拆解" ,
"Agent 工作流设计" ,
"人机确认" ,
"失败恢复" ,
"效果指标拆解" ,
"Badcase 复盘"
]
} ,
{
"name": "技术理解" ,
"items": [
"LLM API" ,
"Tool Use" ,
"Agent Loop" ,
"RAG / Schema Retrieval" ,
"Memory" ,
"Prompt / Context Engineering" ,
"MCP / Skill / Subagent"
]
} ,
{
"name": "方法工具" ,
"items": [
"PRD / Spec" ,
"STAR 需求拆解" ,
"A/B 实验与因果推断" ,
"SQL / 数据分析" ,
"Agent 审计框架"
]
}
] ,
"contact": [ ] ,
"meta": {
"updatedAt": "2026-07-09T07:33:28.080Z"
}
}