职得Offer校园求职助手

提供校园招聘岗位和面试经验查询，支持职位搜索、详情查看及面试复盘信息检索。

Security Scan

The skill's code, runtime instructions, and required secret (a single API key) align with its stated purpose of querying the 职得Offer MCP API; there are no surprising installs, unrelated credentials, or opaque behaviors.

Purpose & Capability

Name/description (校园招聘岗位和面经查询) match the implementation: Node scripts call a single external MCP endpoint (offer.yxzrkj.cn) to run jobs.search/jobs.get/interviews.search/interviews.get. No unrelated services, binaries, or credentials are requested.

Instruction Scope

SKILL.md and the scripts limit actions to reading an API key (env var or scripts/config.json), composing MCP JSON-RPC requests, and printing results. Instructions do not reference reading other system files, sending data to unrelated endpoints, or collecting extra context.

Install Mechanism

There is no install spec or remote download. This is an instruction+script package (local Node scripts). Nothing in the package fetches or installs external code at runtime.

Credentials

Only one credential is required (ZHIDE_OFFER_KEY or scripts/config.json.zhideOfferKey), which is appropriate for authenticating to the described MCP API. No other secrets, config paths, or unrelated env vars are requested.

Persistence & Privilege

Skill does not request always:true or broader platform privileges and does not modify other skills or system-wide settings. It can be invoked by the agent (default), which is expected for a skill of this type.

Assessment

This package is a straightforward local wrapper around the 职得Offer MCP API and needs a single API key (ZHIDE_OFFER_KEY). Before installing/providing a key: verify you trust the endpoint (https://offer.yxzrkj.cn), keep your real key in an environment variable rather than a checked-in scripts/config.json, and avoid committing the config file to public repos. If you have concerns about granting an API key to third-party code, create/restrict a dedicated key that can be revoked.

职得Offer校园求职技能

通过 MCP 协议调用职得Offer接口，查询校招岗位和面试经验。

适合的典型场景：

• 想找某个方向的校招或实习岗位
• 想看某公司某岗位的真实面经
• 想快速做面试前的信息收集

配置

API Key 从配置文件或环境变量读取，不硬编码：

const KEY = process.env.ZHIDE_OFFER_KEY || require('./config.json').zhideOfferKey;

用户需在 scripts/config.json 中配置：

{ "zhideOfferKey": "ofk_你的key" }

MCP 调用规范

• 端点：https://offer.yxzrkj.cn/mcp（POST）
• Headers：Content-Type: application/json、Accept: application/json, text/event-stream、Authorization: Bearer <KEY>
• 每次调用前需先 initialize 握手（无状态，每次独立 POST）
• 脚本已封装，直接调用即可

快速使用

# 查岗位
node ~/.openclaw/skills/zhide-offer/scripts/jobs_search.js 数据产品经理

# 看岗位详情
node ~/.openclaw/skills/zhide-offer/scripts/jobs_get.js <岗位ID>

# 查面经
node ~/.openclaw/skills/zhide-offer/scripts/interviews_search.js 产品经理 --company 字节跳动 --tag 校招 --limit 5

# 看完整面经
node ~/.openclaw/skills/zhide-offer/scripts/interviews_get.js <面经ID>

如果想用统一入口，也可以：

bash ~/.openclaw/skills/zhide-offer/scripts/zhide_offer.sh jobs-search 产品经理 --size 3
bash ~/.openclaw/skills/zhide-offer/scripts/zhide_offer.sh interviews-search 产品经理 --limit 2

工作流

搜索岗位

1. 运行 scripts/jobs_search.js，传入 keyword（可选 company/city/pageSize）
2. 结果在 result.structuredContent.data.items[]，字段见 references/api.md
3. 如需详情，用岗位 id 运行 scripts/jobs_get.js

搜索面经

1. 运行 scripts/interviews_search.js，传入 position_query（精准单词组效果最好，多词组合可能返回空）
2. 结果在 result.structuredContent.data.items[]
3. 如需全文，用面经 id 运行 scripts/interviews_get.js

注意事项

• 发布版本不要携带真实 scripts/config.json，公开包只保留模板文件
• position_query 用单一关键词（如"数据产品经理"），不要堆多个词
• interviews.search 响应较慢（约8秒），正常现象
• 每日调用限额200次（account.entitlements 可查剩余）
• 详细字段说明见 references/api.md