$ loading_
Hybrid BM25 and dense search over a corpus with cited passages and sources.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "insight-mcp" yet — see the docs or source repo.
Search the configured corpus for material about "hybrid retrieval" and return the most relevant passages and sources for citation.
Returns relevant passages with source metadata for easy citation.
For a question about RAG retrieval strategies, search the corpus first and then summarize the evidence for direct use in an answer.
Produces evidence-backed summaries with sources, not just conclusions.
Researchers can quickly find supporting evidence in a configured corpus before writing or analyzing. It returns passages and sources for verification and citation.
When an AI needs to answer from internal documents, it can retrieve relevant content first and include sources in the response. This fits citation-heavy use cases.
Developers can plug it into an MCP workflow to run hybrid BM25 and dense retrieval over a configurable corpus. It suits applications that need both keyword and semantic matching.
It performs hybrid BM25 and dense retrieval over a configurable corpus via MCP, returning relevant passages and sources. It is useful for evidence-backed AI answers.
You need a configured corpus first. Beyond that, the provided information does not specify additional installation or runtime requirements.
It combines keyword retrieval (BM25) and semantic retrieval (dense), making it better at both exact matching and semantic recall. It also returns source passages for citation.
Search documents with hybrid retrieval, reranking, guardrails, and tracing.
Search local documents by keyword, fetch passages, and list sources offline.
Search locally indexed PDF docs with citations and context for LLM agents.
Run hybrid search over self-hosted public data indexes via MCP.
Search, filter, and analyze academic literature across multiple research databases.
Search markdown knowledge bases with hybrid ranking and intelligent reranking.