web-searcher
par kassimisai
Web & Frontend DevelopmentRésumé
Autonomous web research agent that performs multi-step searches, follows links, extracts data, and synthesizes.
Description
Ce que fait ce skill
An autonomous research agent that breaks queries into targeted searches, fetches and reads source pages, cross-references findings, and returns structured reports with citations and confidence levels.
Handles multi-step web research in one pass rather than requiring manual searching, tab-switching, and summarizing across sources.
Quand l'utiliser
- Comparing pricing across SaaS tools
- Researching a company before a job interview
- Finding recent news on a niche technical topic
- Compiling a list of open-source alternatives to a paid product
- Verifying a claim against multiple sources
Exemple de workflow
User asks: 'What are the top open-source vector databases and how do they compare on performance?'
- 1
Break query into targeted searches: 'open-source vector databases 2024', 'pgvector vs Weaviate vs Qdrant benchmark', 'Chroma vs Milvus comparison'
- 2
Run up to 3 searches, assess result quality
- 3
Fetch full content from 2-3 authoritative pages (docs, benchmarks, technical blogs)
- 4
Cross-reference findings, flag contradictions, note consensus on performance metrics
- 5
Synthesize into a comparison table with sources and confidence levels
Structured report with key findings, a comparison table, cited URLs, and a note on which claims have limited sourcing
Installation
clawhub install kassimisai/web-searcherExécute la commande CLI, ou colle l'instruction dans Claude Code, Cursor, Codex ou l'agent de code que tu utilises.