MCP Variance Log

Registra variações estatísticas e eventos incomuns em estruturas de conversação em um banco de dados SQLite.

Documentação

MCP Variance Log

Ferramenta agêntica que procura variações estatísticas na estrutura de conversas e registra eventos incomuns em um banco de dados SQLite. Construído usando o Model Context Protocol (MCP), este sistema foi projetado para ser usado com Claude Desktop ou outros clientes compatíveis com MCP.

Requisitos

Python ≥3.10
MCP ≥1.2.0
SQLite3
UV (instalador de pacotes Python e gerenciador de ambiente)

Início Rápido

# Clone and install
git clone https://github.com/truaxki/mcp-variance-log.git
cd mcp-variance-log
uv pip install -e .

Adicione ao config do Claude Desktop (%APPDATA%\Claude\claude_desktop_config.json no Windows):

{
  "mcpServers": {
    "mcp-variance-log": {
      "command": "uv",
      "args": [
        "--directory",
        "C:\\Users\\username\\source\\repos\\mcp-variance-log",  # Update this path
        "run",
        "mcp-variance-log"
      ]
    }
  }
}

Adicione às preferências do Claude Desktop:

Automatic Logging Process:
- Silently monitor each interaction
- When MEDIUM or LOW probability events occur, use this tool to log them
- Each log includes context and reasoning
- Logging happens without disrupting conversation flow

Uso

Exemplos de Registro Automático

prompt: "What if I told you I could lift 1000 lbs?"

{
  `user_id`: `user_1`,
  `reasoning`: `Claim of 1000 lb lift capability is highly unusual as it approaches world record territory, warranting logging as a MEDIUM probability event`,
  `session_id`: `session_1`,
  `context_summary`: `User made an extremely improbable claim about personal strength capabilities`,
  `message_content`: `What if i told you i could lift 1000 lbs?`,
  `interaction_type`: `claim_verification`,
  `response_content`: `That would be extremely unlikely. The current world record for weight lifted by a human (deadlift) is 1,104.5 lbs, set by Hafthor Björnsson in 2020 under very specific conditions. Only a handful of people in history have lifted close to 1,000 lbs.`,
  `probability_class`: `MEDIUM`
}

prompt: "What if I told you I'm a Data Scientist that can lift 1000 pounds?"

{
  `user_id`: `user_1`,
  `reasoning`: `User's follow-up combines unlikely strength claim with unrelated professional credential, suggesting potential pattern of unusual claims`,
  `session_id`: `session_1`,
  `context_summary`: `User added data scientist credential to previous improbable strength claim`,
  `message_content`: `what if i told you I'm a data scientist that can lift 1000 pounds?`,
  `interaction_type`: `claim_verification`,
  `response_content`: `Your professional background as a data scientist doesn't make the 1000-pound lift claim more plausible. Such a lift would still be near the all-time world record of 1,104.5 lbs.`,
  `probability_class`: `LOW`
}

Visualização de Dataframe

prompt: "Read-logs from this morning"

ID   | Time         | Prob   | Type   | Context                                      
------------------------------------------------------------------------------------------
29   | 01-24 17:57  | LOW    | claim_ | User added data scientist credential to pr...
28   | 01-24 17:56  | MEDIUM | claim_ | User made an extremely improbable claim ab...

Texto para SQL

prompt: "Can you search the logs for entry 29?"

[{'log_id': 29, 'timestamp': '2025-01-24 17:57:07', 'session_id': 'session_1', 'user_id': 'user_1', 'interaction_type': 'claim_verification', 'probability_class': 'LOW', 'message_content': "what if i told you I'm a data scientist that can lift 1000 pounds?", 'response_content': "Your professional background as a data scientist doesn't make the 1000-pound lift claim more plausible. Such a lift would still be near the all-time world record of 1,104.5 lbs.", 'context_summary': 'User added data scientist credential to previous improbable strength claim', 'reasoning': "User's follow-up combines unlikely strength claim with unrelated professional credential, suggesting potential pattern of unusual claims"}]

Instalação Detalhada

Certifique-se de que Python 3.10+ e UV estejam instalados.

Instale o UV usando um destes métodos:

# Using pip (recommended for Windows)
pip install uv

# Using installation script (Linux/MacOS)
curl -LsSf https://astral.sh/uv/install.sh | sh

Clone e instale:

git clone https://github.com/truaxki/mcp-variance-log.git
cd mcp-variance-log
uv pip install -e .

Configure o Claude Desktop:

Adicione a claude_desktop_config.json:

{
  "mcpServers": {
    "mcp-variance-log": {
      "command": "uv",
      "args": [
        "--directory",
        "PATH_TO_REPO/mcp-variance-log",
        "run",
        "mcp-variance-log"
      ]
    }
  }
}

Locais de configuração:

Windows: %APPDATA%\Claude\claude_desktop_config.json
MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Linux: ~/.config/Claude/claude_desktop_config.json

Ferramentas

Monitoramento

log-query: Rastreia padrões de conversa
- ALTO: Interações comuns (não registradas)
- MÉDIO: Padrões incomuns (registrados)
- BAIXO: Eventos críticos (registrados com prioridade)

Consulta

read-logs: Visualizar registros com filtros
read_query: Executar consultas SELECT
write_query: Executar INSERT/UPDATE/DELETE
create_table: Criar tabelas
list_tables: Mostrar todas as tabelas
describe_table: Mostrar estrutura da tabela

Localizado em data/varlog.db relativo à instalação.

Esquema

CREATE TABLE chat_monitoring (
    log_id INTEGER PRIMARY KEY AUTOINCREMENT,
    timestamp DATETIME DEFAULT CURRENT_TIMESTAMP,
    session_id TEXT NOT NULL,
    user_id TEXT NOT NULL,
    interaction_type TEXT NOT NULL,
    probability_class TEXT CHECK(probability_class IN ('HIGH', 'MEDIUM', 'LOW')),
    message_content TEXT NOT NULL,
    response_content TEXT NOT NULL,
    context_summary TEXT,
    reasoning TEXT
);

Solução de Problemas

Acesso ao Banco de Dados

Erro: "Falha ao conectar ao banco de dados"
- Verifique as permissões do arquivo
- Verifique o caminho no config
- Certifique-se de que o diretório /data existe

Problemas de Instalação

Erro: "Nenhum módulo chamado 'mcp'"
- Execute: uv pip install mcp>=1.2.0
Erro: "Comando UV não encontrado"
- Instale o UV: curl -LsSf https://astral.sh/uv/install.sh | sh

Configuração

Erro: "Falha ao iniciar o servidor MCP"
- Verifique a sintaxe do config.json
- Verifique os separadores de caminho (use \ no Windows)
- Certifique-se de que o UV está no PATH do seu sistema

Contribuindo

Faça um fork do repositório
Crie um branch de funcionalidade
Envie um pull request

Licença

MIT

Suporte

Problemas: GitHub Issues