Expert guidelines for Chrome extension development with Manifest V3, covering security, performance, and best practices
日本語の概要は準備中です。原文の説明を表示しています。
Searches multiple web sources, synthesizes findings, and produces cited research reports using delegated subagents. Use when the user asks to research a topic online, search the web, look something up, find current information, compare options, or produce a research report.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Before delegating to subagents, you MUST:
Create a research folder - Organize all research files in a dedicated folder relative to the current working directory:
mkdir research_[topic_name]
This keeps files organized and prevents clutter in the working directory.
Analyze the research question - Break it down into distinct, non-overlapping subtopics
Write a research plan file - Use the write_file tool to create research_[topic_name]/research_plan.md containing:
Planning Guidelines:
For each subtopic in your plan:
Use the task tool to spawn a research subagent with:
research_[topic_name]/findings_[subtopic].mdRun up to 3 subagents in parallel for efficient research
Subagent Instructions Template:
Research [SPECIFIC TOPIC]. Use the web_search tool to gather information.
After completing your research, use write_file to save your findings to research_[topic_name]/findings_[subtopic].md.
Include key facts, relevant quotes, and source URLs.
Use 3-5 web searches maximum.
After all subagents complete:
Review the findings files that were saved locally:
list_files research_[topic_name] to see what files were createdread_file with the file paths (e.g., research_[topic_name]/findings_*.md)read_file for LOCAL files only, not URLsSynthesize the information - Create a comprehensive response that:
Write final report (optional) - Use write_file to create research_[topic_name]/research_report.md if requested
Note: If you need to fetch additional information from URLs, use the fetch_url tool, not read_file.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Expert guidelines for Chrome extension development with Manifest V3, covering security, performance, and best practices
日本語の概要は準備中です。原文の説明を表示しています。
Use this skill when building applications with Gemini models, Gemini API, working with multimodal content (text, images, audio, video), implementing function calling, using structured outputs, or needing current model specifications. Covers SDK usage (google-genai for Python, @google/genai for JavaScript/TypeScript, com.google.genai:google-genai for Java, google.golang.org/genai for Go), model selection, and API capabilities.
日本語の概要は準備中です。原文の説明を表示しています。
ScepticAgent internal architecture reference. Use whenever the user asks how the extension works, wants to add a new AI provider, add a new highlight category, debug communication between components, understand the agent loop or streaming, or work with provider routing and the Gemini CORS proxy.
日本語の概要は準備中です。原文の説明を表示しています。
ScepticAgent AI prompt reference. Use whenever the user asks about the prompts, wants to change analysis behaviour, modify what the AI looks for, adjust highlight categories or scoring, understand what instructions are sent to the AI, or tune the output format of analysis or highlights.
日本語の概要は準備中です。原文の説明を表示しています。
ScepticAgent packaging and Chrome Web Store publishing workflow. Use whenever the user wants to build the zip, package the extension, publish or submit to the Chrome Web Store, update the store listing, or deal with manifest/icon/permission requirements for publishing.
日本語の概要は準備中です。原文の説明を表示しています。
ScepticAgent UI and styles reference. Use whenever the user wants to change the side panel UI, modify highlight colours, update the markdown renderer, work with CSS architecture, add or change category colours, or understand how the sidepanel renders analysis output.
日本語の概要は準備中です。原文の説明を表示しています。