Rivalsearchmcp MCP Server

RivalSearchMCP combines general web discovery with searches tailored to scholarly work, datasets, social discussions, news, and public GitHub repositories. Its web search queries several engines concurrently, merges duplicate results, and allows the remaining engines to continue if one fails. Scientific searches cover sources such as OpenAlex, Crossref, arXiv, and PubMed, while dataset discovery reaches repositories including Kaggle and Zenodo. Community searches include Reddit, Hacker News, and Bluesky; news aggregation removes duplicates using URLs and title similarity.

After discovery, website mapping and content operations support site exploration, URL retrieval, streaming, text analysis, and link extraction. Content operations can also score sources, look for conflicting claims, and check URL liveness, redirects, paywalls, and modification dates. Document analysis extracts text and metadata from PDFs, DOCX files, Markdown, plain text, and images; OCR handles scanned PDFs and image content. The research_topic workflow offers open-ended topic research or a cross-source profile of a named entity. Both modes attach result-level quality scores and an overall confidence signal, giving AI assistants information to weigh when using the collected material.