Overwatch MCP

MCP server for querying Graylog, Prometheus, and InfluxDB 2.x from Claude Desktop.

Tools

Quick Start

One-Line Setup (Docker)

curl -fsSL https://raw.githubusercontent.com/malindarathnayake/Overwatch-mcp/main/compose/setup.sh | bash
cd Overwatch_MCP
# Edit .env and config.yaml with your values
docker compose up -d

Manual Setup (Docker)

# Download compose files
mkdir -p Overwatch_MCP && cd Overwatch_MCP
curl -fsSLO https://raw.githubusercontent.com/malindarathnayake/Overwatch-mcp/main/compose/docker-compose.yml
curl -fsSLO https://raw.githubusercontent.com/malindarathnayake/Overwatch-mcp/main/compose/.env.example
curl -fsSLO https://raw.githubusercontent.com/malindarathnayake/Overwatch-mcp/main/compose/config.example.yaml

# Create config from templates
cp .env.example .env
cp config.example.yaml config.yaml

# Edit .env with your credentials
# Edit config.yaml if needed (adjust allowed_buckets, limits, etc.)

# Run
docker compose up -d

Local Install

pip install -e .
cp .env.example .env
cp config/config.example.yaml config/config.yaml
# Edit both files with your values
python -m overwatch_mcp

Claude Desktop Config

Docker

~/.claude/config.json (Linux/Mac) or %APPDATA%\Claude\config.json (Windows):

{
  "mcpServers": {
    "overwatch": {
      "command": "docker",
      "args": [
        "run", "--rm", "-i",
        "-v", "/path/to/config:/app/config:ro",
        "--env-file", "/path/to/.env",
        "ghcr.io/malindarathnayake/Overwatch-mcp:latest"
      ]
    }
  }
}

Local Python

{
  "mcpServers": {
    "overwatch": {
      "command": "python",
      "args": ["-m", "overwatch_mcp"],
      "env": {
        "GRAYLOG_URL": "https://graylog.internal:9000/api",
        "GRAYLOG_TOKEN": "your-token",
        "PROMETHEUS_URL": "http://prometheus.internal:9090",
        "INFLUXDB_URL": "https://influxdb.internal:8086",
        "INFLUXDB_TOKEN": "your-token",
        "INFLUXDB_ORG": "your-org"
      }
    }
  }
}

Windows PowerShell Setup

One-shot script to configure Claude Desktop on Windows:

# Stop Claude if running
Get-Process -Name "Claude*" -ErrorAction SilentlyContinue | Stop-Process -Force

$config = @'
{
  "mcpServers": {
    "overwatch": {
      "command": "C:/Users/<USERNAME>/AppData/Local/Microsoft/WindowsApps/python3.13.exe",
      "args": ["-m", "overwatch_mcp", "--config", "C:/path/to/Overwatch-mcp/compose/config.yaml"],
      "env": {
        "GRAYLOG_URL": "https://your-graylog-url",
        "GRAYLOG_TOKEN": "<YOUR_GRAYLOG_TOKEN>",
        "PROMETHEUS_URL": "http://your-prometheus-url:9090",
        "INFLUXDB_URL": "https://your-influxdb-url",
        "INFLUXDB_TOKEN": "<YOUR_INFLUXDB_TOKEN>",
        "INFLUXDB_ORG": "<YOUR_INFLUXDB_ORG>",
        "LOG_LEVEL": "debug",
        "LOG_FILE": "C:/path/to/Overwatch-mcp/overwatch.log"
      }
    }
  }
}
'@
[System.IO.File]::WriteAllText("$env:APPDATA\Claude\claude_desktop_config.json", $config)

# Install from source (run from repo root)
cd C:\path\to\Overwatch-mcp
pip install -e .

Note: Replace <USERNAME>, <YOUR_GRAYLOG_TOKEN>, <YOUR_INFLUXDB_TOKEN>, <YOUR_INFLUXDB_ORG>, and paths with your actual values.

Configuration

config.yaml

The config uses ${ENV_VAR} substitution - values come from environment at runtime.

server:
  log_level: "info"

datasources:
  graylog:
    enabled: true
    url: "${GRAYLOG_URL}"
    token: "${GRAYLOG_TOKEN}"
    timeout_seconds: 30
    max_time_range_hours: 24
    max_results: 1000
    # Production environments to filter on (auto-builds from known_applications.json)
    production_environments:
      - "prod"
      - "production"
    # Known apps file - auto-builds env filter from discovered data
    known_applications_file: "${GRAYLOG_KNOWN_APPS_FILE:-}"

  prometheus:
    enabled: true
    url: "${PROMETHEUS_URL}"
    timeout_seconds: 30
    max_range_hours: 168

  influxdb:
    enabled: true
    url: "${INFLUXDB_URL}"
    token: "${INFLUXDB_TOKEN}"
    org: "${INFLUXDB_ORG}"
    timeout_seconds: 60
    allowed_buckets:
      - "telegraf"
      - "app_metrics"

cache:
  enabled: true
  default_ttl_seconds: 60

Disable a datasource by setting enabled: false. Server runs in degraded mode if some datasources fail health checks.

Tool Parameters

graylog_search

{
  "query": "level:ERROR AND service:api",
  "from_time": "-2h",
  "to_time": "now",
  "limit": 100,
  "fields": ["timestamp", "message", "level"]
}

Time formats: ISO8601 (2025-01-27T10:00:00Z), relative (-1h, -30m), now

graylog_fields

{
  "pattern": "http_.*",
  "limit": 100
}

prometheus_query

{
  "query": "rate(http_requests_total[5m])",
  "time": "-1h"
}

prometheus_query_range

{
  "query": "up",
  "start": "-6h",
  "end": "now",
  "step": "1m"
}

Step auto-calculated if omitted.

prometheus_metrics

{
  "pattern": "http_.*",
  "limit": 100
}

influxdb_query

{
  "query": "from(bucket: \"telegraf\") |> range(start: -1h) |> filter(fn: (r) => r._measurement == \"cpu\")",
  "bucket": "telegraf"
}

Bucket must be in allowed_buckets config.

Error Codes

Application Discovery

Generate a known applications file to speed up lookups:

# Using environment variables
python scripts/discover_applications.py --env

# Or with explicit credentials
python scripts/discover_applications.py \
  --url https://graylog.example.com \
  --token YOUR_TOKEN \
  --hours 24 \
  --environment "environment:prod" \
  --output known_applications.json

Output known_applications.json:

{
  "_metadata": {
    "generated_at": "2025-01-28T10:00:00",
    "identifier_fields_used": ["application", "service", "container_name"]
  },
  "environments": ["prod", "staging", "dev"],
  "applications": [
    {
      "name": "api-gateway",
      "identifier_fields": ["service", "application"],
      "aliases": [],
      "description": "",
      "team": "",
      "enabled": true
    }
  ]
}

Edit the file to:

• Remove entries you don't need (enabled: false)
• Add descriptions and team ownership
• Add aliases for alternative names

Then set GRAYLOG_KNOWN_APPS_FILE=/path/to/known_applications.json in your environment.

Development

# Install with dev deps
pip install -e ".[dev]"

# Tests
pytest tests/ -v

# Coverage
pytest tests/ -v --cov=overwatch_mcp

Project Structure

src/overwatch_mcp/
├── __main__.py        # Entry point
├── server.py          # MCP server
├── config.py          # Config loader
├── cache.py           # TTL cache
├── clients/           # HTTP clients (graylog, prometheus, influxdb)
├── tools/             # MCP tool implementations
└── models/            # Pydantic models

127 tests (89 unit, 38 integration).

Usage Guide

See Docs/usage-guide.md for examples of how to ask questions:

• Finding errors and investigating issues
• Searching logs with filters and time ranges
• Querying metrics and trends
• Investigation workflows and common patterns

Troubleshooting

Server won't start: Check config/config.yaml exists and env vars are set.

Datasource unavailable: Verify URL, check token permissions. Server continues with available datasources.

Query errors: Check syntax (Lucene/PromQL/Flux), verify time range within limits, ensure bucket is allowlisted for InfluxDB.

License

MIT