mcp-documentacao
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-documentacaosearch for documentation on chunking parameters"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-documentacao
MCP server que indexa documentos técnicos e disponibiliza busca via RAG híbrida (vetorial + BM25). Apenas recuperação — sem geração.
Uso
pip install .
python3 src/server.pyO servidor roda via stdio — compatível com qualquer cliente MCP.
Related MCP server: RAG-MCP
Configuração
Edite config.yaml:
sources:
- path: ./src/guides
type: documento
description: Documentos técnicos
db_path: ./chroma_db
collection_name: documents
embedding:
provider: huggingface # huggingface | openai | jina | ollama
model: paraphrase-multilingual-MiniLM-L12-v2
search:
top_k: 5
hybrid: true
cache_max_size: 200
cache_ttl: 3600
chunking:
max_size: 1000
overlap: 200CONFIG_PATH pode ser passada como variável de ambiente.
Ferramentas MCP
Ferramenta | Descrição |
| Busca trechos relevantes |
| Lista documentos indexados |
| Reindexa todas as fontes |
O parâmetro fonte filtra por tipo de fonte (ex: documento, faq).
Fontes
Múltiplas fontes com tipos arbitrários. Formatos suportados: Markdown (.md) e texto (.txt). Frontmatter YAML é lido como metadados (descrição, palavras-chave).
Docker
docker compose up --buildMonte volumes para /app/src/guides e /app/chroma_db para persistência.
Testes
pip install pytest
pytest tests/Provedores de embedding
huggingface (padrão, local) —
all-MiniLM-L6-v2,paraphrase-multilingual-MiniLM-L12-v2openai —
text-embedding-3-small,text-embedding-3-large(requerOPENAI_API_KEY)jina —
jina-embeddings-v3(requerJINA_API_KEY)ollama — modelos locais via Ollama
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for querying Forkast documentation
Hybrid human + AI expertise for faster, trusted answers and decisions via MCP Server.
Token-free MCP server for structured RevoGrid Core, Pro, and Enterprise knowledge retrieval.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceAn MCP server that indexes documents and serves relevant context to LLMs via Retrieval Augmented Generation (RAG).15 npm37MIT
- FlicenseNot gradedqualityDmaintenanceA Retrieval Augmented Generation MCP server that ingests documents into a local vector database and enables semantic search queries.9-
- AlicenseNot gradedqualityDmaintenanceMCP server for documentation search that automatically indexes web documentation sites and provides semantic, full-text, or hybrid search capabilities.4 npmMIT
- FlicenseNot gradedqualityDmaintenanceRead-only MCP server with hybrid search combining dense semantic and sparse keyword retrieval via Qdrant, enabling document querying and fetching for ChatGPT Deep Research.-