Knowledge Raven

Make your knowledge agent-ready.

CommunityOnlineNo sign-inGlobalFreeRead-only

What it can do

  • List Knowledge Bases: Lists all available knowledge bases with document counts.
  • Search Knowledge Base: Targeted search for specific facts, definitions, or narrow questions. Returns up to 10 best matches with context. Each result includes a source_link for deep-linking to the cite
  • Broad Search: Keyword-heavy search for overview, comparison, and exploratory questions ('What tools do we use?', 'List all...'). Returns up to 30 results with broader matching. Each result includes a

What data it sees

Do you need an account

No: the server works without sign-in

Make your knowledge agent-ready. Connect docs from Confluence, Notion, GitHub, Dropbox, or Google Drive — any AI agent searches them via one MCP endpoint. 3 retrieval modes: vector search, broad search, and full document access.

Server tool list (6)

Raw names from tools/list. Only developers need these.

list_knowledge_basesLists all available knowledge bases with document counts.
search_knowledge_baseTargeted search for specific facts, definitions, or narrow questions. Returns up to 10 best matches with context. Each result includes a source_link for deep-linking to the cited passage. For overview or aggregation questions, use broad_search instead.
broad_searchKeyword-heavy search for overview, comparison, and exploratory questions ('What tools do we use?', 'List all...'). Returns up to 30 results with broader matching. Each result includes a source_link for deep-linking to the cited passage. For specific facts, use search_knowledge_base.
list_documentsLists all documents in a knowledge base with titles, types, AI-generated summaries, and source_link for each document. Use this to understand what content exists before searching, or when the user asks about available documents.
fetch_documentRetrieves document content flexibly. Modes: 'preview' (first 3 sections + summary), 'full' (entire document), 'chunks' (specific sections by index range, e.g. chunk_range=[5,10]). Returns a chunks array with per-chunk source_link for deep-linking to specific passages, plus a document-level source_link. Start with preview to check relevance before fetching full content.
get_document_metadataRetrieves metadata for a document (title, type, size, page count, etc.) including a source_link to view the document on the platform.