Web Search Enhanced

Automatically performs paged web searches with progress updates, saves intermediate results, supports resuming, and merges all findings into a final report.

Security Scan

The skill claims to perform paged web searches and store intermediate results, but the shipped code does not perform real web searches and has mismatches in how it stores/reads intermediate data—this inconsistency could be benign (unfinished or mock implementation) but is misleading and warrants caution.

Purpose & Capability

The skill's name/description promise automated web searches and merged results from many pages, but the implementation's searchQuery() creates simulated results (no network/search engine calls). A user expecting real web search functionality would not get it. The SKILL.md says it uses a 'Memory system' for intermediate results, but the code writes a local file (.web-search-temp-results.md) instead—mismatch between claimed capability and actual code.

Instruction Scope

SKILL.md describes saving intermediate results and resuming, which the code attempts, but the instructions claim use of a Memory system and real web searching while the code: (a) simulates results locally, (b) writes/reads a local temp file, and (c) uses webFetch() to read that local file (an odd choice). The instructions are high-level and leave implementation details vague, granting broad discretion that produced this mismatch.

Install Mechanism

No install spec is provided (instruction-only style plus one code file). Nothing is downloaded or installed at runtime, so install mechanism risk is low.

Credentials

The skill requests no environment variables, credentials, or config paths. The level of requested access is proportionate to the visible behavior (it only writes a local temp file).

Persistence & Privilege

always is false and the skill does not modify other skills or system-wide configs. It creates a local temp file (.web-search-temp-results.md) in the working directory and removes it on completion—this is modest persistence and within expected scope.

What to consider before installing

This skill is misleading: it claims to do paged web searches but the code only returns simulated (mock) search results and uses a local temp file (.web-search-temp-results.md). Before installing or relying on it: (1) Do not assume it performs real web searches—test it in a sandbox to confirm behavior. (2) If you need real web search, ask the author or check the code to see which search API it should call (and where credentials would be provided). (3) Note it will create a file named .web-search-temp-results.md in the agent's working directory; ensure that location is acceptable and contains no sensitive data. (4) The use of webFetch() to read a local file is suspicious/incorrect—verify how webFetch is implemented in your environment to ensure it won't make unexpected network requests. If you want to proceed only for development/testing, do so in an isolated environment and request a corrected implementation from the author that actually integrates with a search API and uses a clear, safe storage mechanism.

Web Search Enhanced - 增强版网页搜索

功能

这个技能解决了普通 Web Search 的文本长度限制问题：

• ✅ 自动分页搜索 - 当搜索结果过多时自动分批处理
• ✅ 中间结果保存 - 每次搜索完成后保存到临时文件，不会丢失
• ✅ 自动合并结果 - 最后一次性返回所有搜索结果
• ✅ 进度提示 - 实时告诉用户当前进度
• ✅ 断点续搜 - 如果中途中断，可以继续搜索剩余部分

使用场景

• 搜索大量网页（超过 10 个结果）
• 需要汇总多个来源的信息
• 长文档/长文章搜索
• 需要对比多个网页内容

使用方式

直接说：

帮我查找 [搜索内容]

技能会自动判断是否需要分页搜索。

技术实现

• 使用 Memory 系统存储中间结果
• 自动合并搜索结果
• 支持断点续搜