Knowledge RAG MCP Server

by lyonzin• 299•• Scanned•Claim this listing•Request removal
Data Science & MLantigravityclaudeclaude codeclaude code clicodex

Local RAG MCP server for Claude Code — hybrid search (semantic + BM25), cross-encoder reranking, 13 MCP tools, 20 format parsers. Zero external servers, zero API keys.

Knowledge RAG is an open-source mcp server by lyonzin, written mainly in Python, with 299 GitHub stars and the MIT license. Its code was last updated Oct 25, 2026. The latest npm release is knowledge-rag 5.1.0, published Nov 14, 2026, with 421 downloads last week.

Scan summary: scanned Nov 5, 2026 at commit 2be09ec. Seen able to run dynamic code. Scan policy

aboutREADMEsecurityreviews

How to install Knowledge RAG

Built from the start command in the project's README: npx -y knowledge-rag. Check the README for required keys or settings.

Claude Code · Terminal
claude mcp add knowledge-rag -- npx -y knowledge-rag
Claude Desktop · claude_desktop_config.json
{
  "mcpServers": {
    "knowledge-rag": {
      "command": "npx",
      "args": [
        "-y",
        "knowledge-rag"
      ]
    }
  }
}
Cursor · .cursor/mcp.json
{
  "mcpServers": {
    "knowledge-rag": {
      "command": "npx",
      "args": [
        "-y",
        "knowledge-rag"
      ]
    }
  }
}
VS Code · Terminal
code --add-mcp '{"name":"knowledge-rag","command":"npx","args":["-y","knowledge-rag"]}'

What Knowledge Rag can touch

  • Run shell commandsNot seen in this scan
  • Run dynamic codeListed as a tool: evaluate_retrieval
  • Read or write filesNot seen in this scan
  • Read environment and secretsNot seen in this scan
  • Control a web browserNot seen in this scan
  • Connect to databasesLikely, from its own description
  • Read or send messagesNot seen in this scan
  • Touch money or tradingNot seen in this scan
  • Connect to the internetNo connections seen while running

"Not seen" means our checks did not find it, and the tool may still do it. Covers commit 2be09ec, Nov 5, 2026. See the scan policy.

Before you install Knowledge Rag

AgenticMart scanned Knowledge Rag on Nov 5, 2026 with automated checks for hidden code, known-malicious packages, leaked secrets and vulnerable dependencies. The result covers one commit under our scan policy, so review what access you give it and check any advisories shown here.

What it can access

  • An MCP server is a program your AI app starts and talks to. A local one runs with your user's permissions: it can see the files, environment variables and network your account can.
  • Based on its description and topics, it likely needs to connect to a database.

Before you install

  1. Give it a read-only database user unless it must write.
  2. Pin the version you reviewed instead of always installing the latest release.
  3. Read the tool list or scripts before the first run, and keep approval prompts on for actions that write, send or delete.
  • Maintained by lyonzin on GitHub, 299 stars and 49 forks. Popularity is not proof of safety.
  • Last code change: Oct 25, 2026.
  • Open-source license: MIT.
Read the scan reportWhat our scan checks

Tools Knowledge Rag exposes (13)

Listed by the server itself during our scan of commit 2be09ec.

  • search_knowledge

    Hybrid search combining semantic search + BM25 keyword search with cross-encoder reranking. Read-only. No side effects. Args: query: Search query text (1–3 keywords recommended; phrase queries also work) max_results: Maximum number of results (default: 5, capped by sea

  • get_document

    Get the full content of a specific document by filepath. Read-only. No side effects. Args: filepath: Relative path to the document within the documents directory (e.g., "security/technique.md"). Must be an indexed file — use search_knowledge() to find t

  • reindex_documents

    Index or reindex all documents in the knowledge base (runs in background). No flags — incremental pass: index new or changed files, skip unchanged ones, prune deleted sources. Use after filesystem edits outside add_document/update_document. ``force`` — smart reindex: re-embed every

  • get_reindex_status

    Get the current status of a background reindex operation. Lightweight — does not compute full index statistics. Use this to poll progress after calling reindex_documents(). Returns: JSON string with reindex status. When active: operation name, progress (processed/total),

  • list_categories

    List all document categories with their document counts. Read-only. No side effects. Reflects the live index state. Returns: JSON string with category names, document counts per category, and total document count. Usage: Use before filtering search_knowledge() or list_doc

  • list_documents

    List indexed documents one page at a time, optionally filtered by category. Read-only. No side effects. Args: category: Optional category filter. Must be a valid category name — call list_categories() to see available options (e.g., security, ctf, logscale,

  • get_index_stats

    Get statistics and health metrics for the knowledge base index. Read-only. No side effects. Returns: JSON string with system metrics: total documents, total chunks, embedding model name, BM25 status, query cache hit rate, and file watcher status. Usage: Use for sy

  • add_document

    Add a new document to the knowledge base from raw text content. Mutating — writes a file to disk and indexes it immediately. No auth required. Args: content: Full text content of the document (markdown supported) filepath: Relative path within documents directory (e.g.

  • update_document

    Update the content of an existing document in the knowledge base. Mutating — overwrites the file on disk and re-indexes immediately. Old chunks are removed and replaced with new ones. Full content replacement, not a patch. Args: filepath: Full or relative path to the docum

  • remove_document

    Remove a document from the knowledge base index. Mutating — removes index entries. If delete_file=True, also permanently deletes the file from disk (irreversible, cannot be undone). Args: filepath: Path to the document file. Must be an indexed document — list_d

  • add_from_url

    Fetch content from a URL, convert to markdown, and add to the knowledge base. Mutating — makes an outbound HTTP request (requires internet access), strips HTML, converts to markdown, saves to disk, and indexes immediately. Args: url: Full URL to fetch (https:// required).

  • search_similar

    Find documents semantically similar to a given reference document. Read-only. No side effects. Uses the document's embedding for similarity comparison. Args: filepath: Path to the reference document (must already be indexed — list_documents() lists indexed paths, o

  • evaluate_retrieval

    Evaluate search quality by testing whether search_knowledge() retrieves expected documents. Read-only. Runs multiple search queries internally. No side effects on the index. Args: test_cases: JSON string array of test cases. Each item requires "query" (search string)

Knowledge Rag is an open-source mcp server by lyonzin, written mainly in Python. Local RAG MCP server for Claude Code — hybrid search (semantic + BM25), cross-encoder reranking, 13 MCP tools, 20 format parsers. Zero external servers, zero API keys.

At a glance

Stars
299
Forks
49
Language
Python
License
MIT
Created
Feb 2026
Last update
Oct 25, 2026
Type
MCP Server
Price
Free

Topics

antigravityclaudeclaude-codeclaude-code-clicodexcursor-aidocument-searchhybrid-searchinteligencia-artificialknowledge-baselocal-aimcpmcp-serverragrag-chatbotrag-pipelinererankingretrieval-augmented-generationsemantic-searchvector-database

Questions about Knowledge Rag

What is Knowledge Rag?

Local RAG MCP server for Claude Code — hybrid search (semantic + BM25), cross-encoder reranking, 13 MCP tools, 20 format parsers. Zero external servers, zero API keys.

Who maintains Knowledge Rag?

Knowledge Rag is maintained by lyonzin on GitHub (lyonzin/knowledge-rag). AgenticMart lists it from public repository data.

Is Knowledge Rag safe to use?

AgenticMart scanned Knowledge Rag on Nov 5, 2026 with automated checks for hidden code, known-malicious packages, leaked secrets and vulnerable dependencies. The result covers one commit under our scan policy, so review what access you give it and check any advisories shown here.

What license does Knowledge Rag use?

It is published under the MIT license.

Knowledge RAG alternatives

View more →

Similar mcp servers in Data Science & ML, ranked by GitHub stars. Compare licence, recent activity and scan status before you choose.

NameStarsLicenceLast updatedScan
Knowledge RAG (this page)
lyonzin
299MITOct 25, 2026Scanned, no flagged issues
MaxKB
1Panel-dev
23kGPL-3.0Nov 1, 2026Scanned, no flagged issues
FunASR
modelscope
21kMITNov 1, 2026Scanned, no flagged issues
Ida Pro MCP
mrexodia
13kMITSep 26, 2026Scanned, no flagged issues
Pal MCP Server
BeehiveInnovations
12kNone foundDec 15, 2025Not scanned yet
Osaurus
osaurus-ai
8.1kMITNov 1, 2026Scanned, no flagged issues