searching-assistant
par urrrich0
Git & GitHubRésumé
You are the leader of searching group (搜索组组长)
Description
Ce que fait ce skill
Orchestrates multi-agent search by decomposing a query into up to 8 independent sub-tasks, then routing each to the most suitable specialized agent in parallel. Always invokes General_Search_Agent and selects additional agents based on query type. Excludes Academic_Search for date-specific queries.
Running sub-tasks in parallel across specialized agents returns broader coverage faster than single sequential searches.
Quand l'utiliser
- Researching a topic that spans news, web, and academic sources simultaneously
- Finding recent product announcements without triggering academic-only databases
- Answering multi-part questions that need different search strategies
- Gathering background on a company or person from multiple angles
- Compiling a competitive analysis split into parallel sub-queries
Exemple de workflow
User asks: What are the latest breakthroughs in solid-state batteries and who are the leading companies?
- 1
Breaks query into sub-tasks: recent news, company landscape, technical developments
- 2
Assigns each sub-task to the most suitable agent in parallel
- 3
Always routes at least one sub-task to General_Search_Agent
- 4
Collects and merges results from all agents in a single round
Unified summary covering recent breakthroughs, key players, and technical context sourced from multiple specialized agents
Installation
clawhub install urrrich0/searching-assistantExécute la commande CLI, ou colle l'instruction dans Claude Code, Cursor, Codex ou l'agent de code que tu utilises.