NebulaGraph-Skill

A Claude Code skill that enables intelligent interaction with NebulaGraph databases through the nebulagraph-mcp-server. It provides nGQL query generation, schema design guidance, and workflow orchestration for graph database operations.

nebulagraph/
  SKILL.md              # Entry point — workflow decision tree, MCP tool reference, quick patterns
  nGQL-reference.md     # Complete nGQL syntax reference — types, DDL/DML, MATCH, GO, functions, indexes
  data-modeling.md      # Schema design guide — VID strategy, tags, edges, partitions, patterns
  examples.md           # Scenario-based examples — exploration, query composition, error diagnosis

1. A running NebulaGraph cluster (v3.x)
2. nebulagraph-mcp-server configured in your Claude Code MCP settings

Add to your Claude Code MCP config (~/.claude/settings.json or project .claude/settings.json):

{
  "mcpServers": {
    "nebulagraph": {
      "command": "python",
      "args": ["-m", "nebulagraph_mcp_server"],
      "env": {
        "NEBULA_ADDRESS": "127.0.0.1:9669",
        "NEBULA_USER": "root",
        "NEBULA_PASSWORD": "nebula"
      }
    }
  }
}

Copy the nebulagraph/ directory to one of these locations:

Global (available in all projects):

~/.claude/skills/nebulagraph/

Project-level (available only in this project):

<project-root>/.claude/skills/nebulagraph/

1. Start a new Claude Code session
2. Try a trigger phrase like "Query NebulaGraph for all spaces" or "Design a graph schema for user relationships"
3. Claude should activate the skill and use the MCP tools

The skill activates when you mention:

• NebulaGraph, nGQL, graph database schema
• Graph space, graph paths, graph neighbors
• Knowledge graph, graph data modeling
• MCP tool names: list_spaces, `get_s