SolarWinds Logs MCP Server
# SolarWinds Logs MCP Server
A Model Context Protocol (MCP) server for accessing and visualizing SolarWinds Observability logs.
## Note -
This server is currently incomplete as it does not support
structured data search (a limitation of the REST API?). I'm
uncertain if it also needs to accept a data center to use
in the api endpoint calls. Will address both when time allows
(needed it for a real work problem, have to fix that first)
### Tools
#### search_logs
Search SolarWinds Observability logs with optional filtering
- Takes search parameters including filter, time range, and pagination options
- Returns formatted log entries with timestamps, hostnames, and messages
- Supports advanced filtering by group, entity, and more
- Default search range is the last 24 hours
#### visualize_logs
Generate a histogram json response for of log events
- Formatted for Claude and canvas representations
- Configurable time intervals (minute, hour, day)
- Supports UTC or local time zones
- Customizable query filters and time ranges
- Default visualization range is the last 24 hours
### Resources
#### SolarWinds Log Search
- URI Template: `solarwinds://{query}/search`
- Returns log entries matching the specified query
- Example: `solarwinds://error/search`
## Installation
Optionally install from npm:
```bash
npm install -g mcp-solarwinds
```
Or clone and build from source:
```bash
git clone https://github.com/@jakenuts/mcp-solarwinds.git
cd mcp-solarwinds
npm install
npm run build
```
Or just use npx in your configurations
### For Cline VSCode Extension
Add to `%APPDATA%/Code - Insiders/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json`:
```json
{
"mcpServers": {
"solarwinds": {
"command": "npx",
"args": ["-y", "mcp-solarwinds"],
"env": {
"SOLARWINDS_API_TOKEN": "your-api-token"
},
"autoApprove": ["search_logs", "visualize_logs"]
}
}
}
```
### For Claude Desktop
Add to the appropriate config file:
Windows: `%APPDATA%/Claude/claude_desktop_config.json`
MacOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
```json
{
"mcpServers": {
"solarwinds": {
"command": "npx",
"args": ["-y", "mcp-solarwinds"],
"env": {
"SOLARWINDS_API_TOKEN": "your-api-token"
}
}
}
}
```
### Special Windows Configuration
If you encounter the ENOENT spawn npx issue on Windows, use this alternative configuration that specifies the full paths:
```json
{
"mcpServers": {
"solarwinds": {
"command": "C:\\Users\\[username]\\AppData\\Roaming\\nvm\\[node-version]\\node.exe",
"args": [
"C:\\Users\\[username]\\AppData\\Roaming\\npm\\node_modules\\npm\\bin\\npx-cli.js",
"-y",
"mcp-solarwinds"
],
"env": {
"SOLARWINDS_API_TOKEN": "your-api-token"
}
}
}
}
```
## Configuration
The SolarWinds Observability MCP server requires an API token to authenticate with the SolarWinds Observability API.
### Configuration Methods
There are multiple ways to provide the API token:
1. **MCP Settings Configuration (Recommended)**: Configure the token in your MCP settings file
2. **Environment Variable**: Set the `SOLARWINDS_API_TOKEN` environment variable
3. **Local .env File (For Testing)**: Create a `.env` file in the project root with `SOLARWINDS_API_TOKEN=your-token`
For local testing, you can:
1. Copy `.env.example` to `.env` and add your token
2. Run the example script: `node examples/local-test.js`
## Tool Usage Examples
### search_logs
Basic search:
```json
{
"filter": "error"
}
```
Advanced search with time range and pagination:
```json
{
"filter": "error",
"entityId": "web-server",
"startTime": "2025-03-01T00:00:00Z",
"endTime": "2025-03-05T23:59:59Z",
"pageSize": 100,
"direction": "backward"
}
```
### visualize_logs
Basic histogram (ASCII chart):
```json
{
"filter": "error",
"interval": "hour"
}
```
Advanced visualization (ASCII chart):
```json
{
"filter": "error",
"entityId": "web-server",
"startTime": "2025-03-01T00:00:00Z",
"endTime": "2025-03-05T23:59:59Z",
"interval": "day",
"use_utc": true
}
```
Claude visualization (JSON format):
```json
{
"filter": "error",
"interval": "hour",
"format": "json"
}
```
The JSON format returns data that Claude can visualize as a chart:
```json
{
"timeRanges": ["12:02", "12:03", "12:04", "12:05", "12:06", "12:07", "12:08", "12:09"],
"counts": [261, 47, 48, 48, 31, 262, 270, 33],
"total": 1000,
"queryParams": {
"query": "error",
"startTime": "2025-03-05T00:00:00.000Z",
"endTime": "2025-03-05T23:59:59.000Z"
}
}
```
## Development
Install dependencies:
```bash
npm install
```
Build the server:
```bash
npm run build
```
### Debugging
Since MCP servers communicate over stdio, debugging can be challenging. The MCP Inspector provides helpful debugging tools:
```bash
npm run debug:inspector
```
This will provide a URL to access the inspector in your browser, where you can:
- View all MCP messages
- Inspect request/response payloads
- Test tools interactively
- Monitor server state
For local testing without the MCP framework:
```bash
# Create a .env file with your token
cp .env.example .env
# Edit .env to add your token
# Run the example script
node examples/local-test.js
```
## Technical Details
- Built with TypeScript and the MCP SDK
- Uses axios for API communication
- Supports ISO 8601 date formats for time ranges
- Generates ASCII histograms for log visualization
- Default search range: last 24 hours
- Default page size: 50 logs
- Supports multiple authentication methods
TDQS
Scored across 2 tools
The two tools have clearly distinct purposes: one searches logs with filtering, and the other generates visualizations. There is no overlap in functionality, making it impossible to confuse them.
Both tools follow a consistent verb_noun pattern (search_logs, visualize_logs) with the same naming style. This predictability helps agents understand the tool set structure easily.
With only two tools, the server feels too thin for a logs domain, lacking essential operations like creating, updating, or deleting logs. This minimal set limits agent capabilities significantly.
The tool set is severely incomplete for logs management, missing core CRUD operations and lifecycle coverage. Agents can only search and visualize, with no ability to modify or manage log data, leading to dead ends in workflows.