MCP Client Gdelt — AI skill for Claude Code
Agent client for on-demand GraphRAG over geopolitical knowledge graphs — Agent 1 orchestrates 3 MCP servers (GDELT/REBEL/Embedding) to build the graph, Agent 2 (two implementations: Anthropic SDK tool.
How to install MCP Client Gdelt
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open HugoStevenPoveda/MCP-Client-GDELT and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What MCP Client Gdelt does
Agent client for on-demand GraphRAG over geopolitical knowledge graphs — Agent 1 orchestrates 3 MCP servers (GDELT/REBEL/Embedding) to build the graph, Agent 2 (two implementations: Anthropic SDK tool-calling loop and a Claude Code CLI batch runner) retrieves and synthesizes answers from it.
Alternatives in AI
- Wiki Graph — Build the LLM Wiki knowledge graph 2.3k ★
- Anthropic SDK TypeScript — Access to Anthropic's safety-first language model APIs in TypeScript 2.1k ★
- Iwe — Markdown knowledge graph — LSP for your editor, CLI + MCP memory for your AI agents 1.8k ★
README
MCP-Cliente
  
Cliente/agente que consume los 3 servidores [MCP](https://modelcontextprotocol.io) de [`MCP-Server`](../MCP-Server) (GDELT, REBEL, Embedding) para construir un grafo de conocimiento geopolítico bajo demanda y responder consultas sobre él.
Dos agentes
| Agente | Rol | Dónde vive |
|---|---|---|
| Agente 1 — Graph Builder | Orquesta los MCP de GDELT/REBEL/Embedding para construir el grafo on-demand en Neo4j (ingesta → extracción de tripletas → entrenamiento KGE) | `cliente_mcp/` — ConversationOrchestrator + handlers/ |
| Agente 2 — GraphRAG Retrieval | ReAct loop (tool-calling) que consulta el grafo ya construido (BFS, inferencia simbólica, link prediction KGE) y sintetiza la respuesta final en JSON | Dos implementaciones, ver abajo |
El Agente 2 tiene **dos implementaciones** del mismo flujo de retrieval:
cliente_mcp/src/agentGraphRag/agente_sdk.py(condicion_evaluada: "sdk_loop") — loop de tool-calling con el SDK de Anthropic +fastmcp.Client, invocado por el orquestador decliente_mcptras la fase de síntesis dentro de la app Gradio.agente2KGE/(condicion_evaluada: "con_kge") — el mismo flujo pero ejecutado vía Claude Code CLI (CLAUDE.mdcomo prompt del agente +.mcp.jsonapuntando directo alembedding-api), usado para correr un batch de preguntas (preguntas.json→leer_preguntas.py/guardar_resultado.py).
Estructura
MCP-Cliente/
├── cliente_mcp/ # App principal: Gradio UI + Agente 1 + Agente 2 (SDK loop)
│ ├── README.md # Instalación, Docker, variables de entorno
│ └── ARCHITECTURE.md # Diseño detallado de cliente_mcp (patrones, flujo de datos)
└── agente2KGE/ # Runner batch de Agente 2 vía Claude Code
├── CLAUDE.md # Prompt/flujo obligatorio del
Related Skills
LLM Kernel
Rust foundation library for AI-native apps — 16-provider catalog, async LLM client, MCP server, knowledge grap
Octocode
Structural code intelligence for AI agents — semantic search, knowledge graphs, and a built-in MCP server in o
Deeptutor Claude Skill
Claude Code skill adapting DeepTutor's graph-enhanced RAG tutoring methodology. Knowledge graphs, dual-loop so
Claude Agent SDK Laravel
Anthropic Claude Agent SDK for PHP & Laravel — build AI agents with tool use, sandboxing, MCP servers, subagen
Chitta
Persistent memory for AI agents - MCP server + TypeScript SDK for Claude Code, Cursor, Codex & 14 more. 85.8%
Srclight
Deep code indexing MCP server for AI agents. 25 tools: hybrid FTS5 + embedding search, call graphs, git blame/
Related Agents
ML Developer
Machine learning and AI integration specialist. Use for Anthropic SDK, LangChain, LlamaIndex, embedding, vecto
Kg Engineer
Senior Python Engineer for Knowledge Graph integration with CognivAgent. Specializes in Claude Agent SDK MCP t
Canon Architect
Orchestrates Canon automatically — detects ecosystem, closes the knowledge graph, compiles context, runs creat