Manticore Search

io.github.manticoresoftwarePublished by GitHub account manticoresoftwarev0.1.2Updated Oct 24, 2026

MCP server for Manticore Search — lets AI assistants query and manage search databases.

VerifiedSTDIODesktop onlyDatabasesKnowledge & Memory

Overview

AI-generated overview

A local MCP server that lets an AI assistant run SQL queries against a Manticore Search database, inspect its tables, and read Manticore documentation.

What you can do
Run SQL queries against Manticore Search (SELECT, SHOW, DESCRIBE and similar). List tables and indexes, and show a table's schema. List and fetch pages from the Manticore manual. Use a built-in prompt that teaches the assistant Manticore features such as full-text operators, KNN vector search and fuzzy search.
When to use it
Useful when you work with a Manticore Search instance and want an assistant to explore its data, check schemas, or draft and run search queries without leaving the chat.
Requirements
A local Python runtime; the package is installed and run with uvx or pip. A reachable Manticore Search server, configured through MANTICORE_HOST and MANTICORE_PORT (default localhost:9308), with optional MANTICORE_USER and MANTICORE_PASSWORD. Network access to fetch documentation; GITHUB_TOKEN is optional for higher rate limits.
Before you install
Write operations are blocked by default and only enabled with MANTICORE_ALLOW_WRITE_ACCESS; destructive operations such as DROP and TRUNCATE need MANTICORE_ALLOW_DROP. Enabling these lets the assistant change or delete data. Database credentials are passed through environment variables. The optional HTTP transport uses MANTICORE_MCP_AUTH_TOKEN, which should be kept secret.

Installation

In SourceWeft

  1. Open Manticore Search in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

Manticore Search MCP Server

[PyPI - Version] [uv]

MCP server for Manticore Search — enables AI assistants to query and manage Manticore Search databases directly.

Quick Start

Installation

bash
# Option 1: Install with uv (recommended, requires PyPI release)uvx mcp-manticore
# Option 2: Install with pippip install mcp-manticore
# Option 3: Run from source (for local development)uvx --from . mcp-manticore# Or: uv run mcp-manticore

Note:

  • uvx runs the package directly without installation. First-time run may take a moment to download dependencies.
  • The package must be published to PyPI for uvx mcp-manticore to work.
  • For local development or testing unreleased versions, use uvx --from . mcp-manticore

What It Does

Tools

ToolDescription
run_queryExecute SQL queries (SELECT, SHOW, DESCRIBE, etc.)
list_tablesList all tables and indexes
describe_tableGet table schema
list_documentationList available documentation files
get_documentationFetch specific documentation from Manticore manual

Prompts

  • manticore_initial_prompt — Built-in prompt teaching LLMs about Manticore Search features (full-text operators, KNN vector search, fuzzy search, etc.)

Health Check

HTTP endpoint at /health for monitoring connectivity.

Configuration

Environment Variables

VariableDefaultDescription
MANTICORE_HOSTlocalhostManticore server host
MANTICORE_PORT9308HTTP API port
MANTICORE_USER—Username (optional)
MANTICORE_PASSWORD—Password (optional)
MANTICORE_CONNECT_TIMEOUT30Connection timeout (seconds)
MANTICORE_QUERY_TIMEOUT30Query timeout (seconds)
MANTICORE_ALLOW_WRITE_ACCESSfalseEnable write operations (INSERT, UPDATE, DELETE)
MANTICORE_ALLOW_DROPfalseEnable destructive operations (DROP, TRUNCATE)
GITHUB_TOKEN—GitHub token for higher API rate limit
Safety

By default, all write operations are blocked. To enable:

bash
# Enable writes (INSERT, UPDATE, DELETE)export MANTICORE_ALLOW_WRITE_ACCESS=true
# Enable destructive operations (DROP, TRUNCATE)export MANTICORE_ALLOW_DROP=true

Connect to Your AI Assistant

Claude Code

Open terminal and run:

bash
claude mcp add manticore -- uvx mcp-manticore

Or with environment variables:

bash
claude mcp add manticore -- uvx mcp-manticore -- \  MANTICORE_HOST=localhost \  MANTICORE_PORT=9308

For full configuration, edit ~/.claude/mcp_settings.json:

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}

Restart Claude Code or type /mcp restart to apply changes.

Cursor

Method 1: Via Settings UI

  1. Open Cursor → Settings → Tools & MCP
  2. Click "Add MCP Server"
  3. Enter name: manticore
  4. Command: uvx mcp-manticore

Method 2: Via Config File

Global config (~/.cursor/mcp.json):

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}

Project config (.cursor/mcp.json in your project):

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"]    }  }}
Windsurf

Method 1: Via Cascade UI

  1. Open Windsurf → Cascade panel
  2. Click the MCPs icon (🔨) in the top-right
  3. Click "Add Server"
  4. Enter: uvx mcp-manticore

Method 2: Via Config File

Edit ~/.codeium/windsurf/mcp_config.json:

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}

Or open directly in Windsurf: Cmd/Ctrl + Shift + P → "MCP Configuration Panel"

OpenAI Codex

Run in terminal:

bash
codex mcp add manticore -- uvx mcp-manticore

With environment variables:

bash
codex mcp add manticore \  --env MANTICORE_HOST=localhost \  --env MANTICORE_PORT=9308 \  -- uvx mcp-manticore

Or edit ~/.codex/config.toml directly:

toml
[mcp_servers.manticore]command = "uvx"args = ["mcp-manticore"]env = { MANTICORE_HOST = "localhost", MANTICORE_PORT = "9308" }
Gemini CLI

Edit ~/.gemini/settings.json:

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}
VS Code (GitHub Copilot)

Create .vscode/mcp.json in your workspace:

json
{  "servers": {    "manticore": {      "type": "stdio",      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}
Cline
  1. Open Cline panel in VS Code
  2. Click the MCP Servers icon
  3. Click "Configure" → "Add Server"
  4. Select "Command (stdio)" and enter:
    • Name: manticore
    • Command: uvx mcp-manticore

Or edit the MCP settings file directly (accessible via the MCP Servers icon → "Edit Config"):

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}
Roo Code
  1. Open Roo Code panel in VS Code
  2. Click the MCP Servers icon → "Edit MCP Settings"
  3. Add the server configuration

Or create .roo/mcp.json in your project root:

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}
Claude Desktop (Legacy)

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%/Claude/claude_desktop_config.json (Windows):

json
{  "mcpServers": {    "manticore": {      "command": "uvx",      "args": ["mcp-manticore"],      "env": {        "MANTICORE_HOST": "localhost",        "MANTICORE_PORT": "9308"      }    }  }}

HTTP Transport (Remote MCP)

By default, MCP uses stdio (local). For remote access:

bash
export MANTICORE_MCP_SERVER_TRANSPORT=httpexport MANTICORE_MCP_BIND_PORT=8000export MANTICORE_MCP_AUTH_TOKEN="your-secure-token"
uvx mcp-manticore

Connect via URL:

json
{  "mcpServers": {    "manticore": {      "url": "http://localhost:8000/mcp",      "headers": {        "Authorization": "Bearer your-secure-token"      }    }  }}

Troubleshooting

Install uv (required)

macOS / Linux:

bash
# Via installer script (recommended)curl -LsSf https://astral.sh/uv/install.sh | sh
# Or via Homebrewbrew install uv

Windows:

powershell
# Via PowerShellirm https://astral.sh/uv/install.ps1 | iex
# Or via wingetwinget install astral-sh.uv

Verify installation:

bash
uv --version

MCP server not connecting

  1. Verify Manticore is running: curl http://localhost:9308/health
  2. Check environment variables are set correctly
  3. For Claude Code: restart with /mcp restart

Too many tools loaded

Some agents limit active MCP tools. Remove unused servers or use project-scoped configs.

Development

bash
# Clone and setupgit clone https://github.com/manticoresoftware/mcp-manticore.gitcd mcp-manticore
# Install dependenciesuv sync
# Run locallyuv run mcp-manticore
# Run with custom configMANTICORE_HOST=remote-server MANTICORE_PORT=9308 uv run mcp-manticore
# Run testsuv run pytest
# Build packageuv build
# Publish to PyPIuv publish

Architecture

FilePurpose
mcp_manticore/mcp_env.pyConfiguration management
mcp_manticore/mcp_server.pyMCP server implementation
mcp_manticore/manticore_prompt.pyLLM guidance/prompts
mcp_manticore/docs_fetcher.pyGitHub docs fetcher
mcp_manticore/main.pyCLI entry point

License

Apache-2.0

Source: README.md at commit c835b5c

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.1.2LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts