寻人物助手

帮助用户整理失联人员案件信息、记录线索、生成寻人启事、查看寻人进展，并提供搜索指南与防骗提醒。用于用户明确提到寻人、找失联亲友、生成寻人启事、添加线索、查看案件进展、防诈骗提醒等场景。

Security Scan

The skill is internally coherent: it is a local, file-backed missing-persons helper (no network calls or extra credentials) and its code and docs match that purpose, with minor implementation notes about local storage and runtime assumptions.

Purpose & Capability

Name/description match the code and instructions: the skill provides local case creation, clue recording, poster generation, progress views, guides and scam warnings. It does not request unrelated credentials or external services.

Instruction Scope

SKILL.md instructs using the included Node CLI and explicitly limits scope (no police DB, no face recognition, no network). The runtime instructions and code operate on local JSON files under ~/.openclaw/skills-data/looking-for-someone, which is consistent with the stated boundaries.

Install Mechanism

There is no install spec (instruction-only style) — the package includes Node scripts and a package.json but the skill metadata did not declare a required binary. Running the skill requires a Node runtime (node), which is implied by the CLI and tests but not declared in required binaries. No remote downloads or extraction steps are present.

Credentials

The skill requests no environment variables or external credentials. It stores potentially sensitive personal data (phone, ID number masked, birthdate, etc.) in clear JSON files in the user's home directory and the README explicitly notes lack of field-level encryption — this is proportionate to a local helper but presents privacy risk that the user should understand.

Persistence & Privilege

always is false and the skill does not request elevated privileges or modify other skills. It persists its own data under ~/.openclaw/... which is normal for a local assistant but means data remains on disk until deleted.

Assessment

This skill appears to do what it says: it's a local, Node-based CLI that stores case data and generates posters and warnings. Before installing/running: 1) ensure you have Node installed (the skill assumes node but required-binaries was left empty); 2) be aware it stores personal data in plain JSON under ~/.openclaw/skills-data/looking-for-someone and does not implement field-level encryption — avoid putting full ID numbers, bank details, or verification codes into the data if you are concerned about privacy; 3) review the files if you need assurance there are no outbound network calls (the included code is local-only); 4) back up or securely delete the cases.json file when done; and 5) for any high-risk or legal cases, follow the SKILL.md advice to contact police or official channels rather than relying solely on this tool.

Looking For Someone

把这个 skill 当作本地寻人辅助工具，不要把它表述成警方系统、实名数据库或自动联网搜索服务。

执行原则

• 先确认目标：创建案件、补充线索、生成启事、查看进展，还是获取搜索建议。
• 优先建议用户尽快报警，尤其是涉及未成年人、老人、精神疾病、自伤风险、失联时间较长等高风险场景。
• 不要承诺已接入官方数据库、监控系统或实时社交平台搜索。
• 不要声称已实现加密、图像比对、身份核验等未实现能力。
• 输出寻人启事时，默认谨慎处理住址、身份证号、银行卡号等敏感信息。

可用能力

• 创建和查看本地寻人案件
• 添加线索并给出基础分析
• 生成多平台寻人启事文案
• 查看案件进展和下一步建议
• 提供搜索指南
• 输出防骗提醒

运行方式

优先使用 CLI：

node scripts/cli.js 创建 '<案件JSON>'
node scripts/cli.js 列表
node scripts/cli.js 进展 <案件ID>
node scripts/cli.js 线索 <案件ID> <线索内容>
node scripts/cli.js 启事 <案件ID> [general|wechat|weibo|douyin|official]
node scripts/cli.js 指南
node scripts/cli.js 提醒

建案输入建议

创建案件时，至少包含：

• name
• age
• gender
• lastSeenDate
• lastSeenLocation

可选字段：

• phone
• birthDate
• idNumber
• height
• clothing
• distinguishingFeatures
• circumstances
• possibleDestinations
• familyContacts

字段说明与推荐流程详见 references/fields-and-workflows.md。

数据边界

• 当前数据保存在本地目录 ~/.openclaw/skills-data/looking-for-someone/
• 当前版本使用本地 JSON 文件存储案件
• 当前版本未实现字段级加密、图片识别、照片比对、联网抓取、实名核验
• 因此输出时必须避免夸大能力，并提醒用户通过警方和官方渠道核验信息
• 隐私与安全边界详见 references/privacy-and-boundaries.md

结果表述要求

• 把建议表述为“本地整理结果”或“规则型建议”
• 不要把线索分析说成事实判断
• 对高风险案件，优先建议报警和联系官方机构
• 若数据不完整，先引导补充关键信息再生成启事

安全要求

遇到以下情况时，优先提醒风险：

• 有人要求付费才提供线索
• 对方要求先转账、先打保证金、先捐款
• 对方索要身份证、银行卡、验证码
• 对方声称是警方却无法提供可核验身份
• 对方要求下载不明 App、打开屏幕共享、点击陌生链接

必要时直接运行：

node scripts/cli.js 提醒

测试

运行最小测试：

node test.js

测试应覆盖：

• 创建案件
• 查看列表
• 添加线索
• 生成启事
• 查看进展
• 防骗提醒