AWS CloudWatch Log Analyzer MCP lets assistants search log groups, identify error patterns and correlate events using an AWS credential profile.
AWS CloudWatch Log Analyzer MCP lets assistants search log groups, identify error patterns and correlate events using an AWS credential profile.
This AWS Labs example brings CloudWatch log discovery, search and analysis into an MCP client. Tools can search a single group or several groups, filter events, summarize activity, identify repeated errors and correlate matching events. Resources expose additional log-group and stream information. The integration retrieves diagnostic evidence; it does not deploy a fix to the application producing the logs.
The project supplies a uvx configuration that installs its server from the GitHub repository and launches cw-mcp-server over stdio. It supports an AWS profile and region through launch arguments, and can use the normal AWS credential chain. A named profile avoids treating a fixed set of environment variables as the only way to authenticate.
Choose the account, region, log group and time range before searching. A multi-group query is useful when a request crosses services, but matching timestamps alone do not establish the cause of an error. Retain the original messages and request identifiers so the assistant's proposed explanation can be checked.
Find an error visible in the CloudWatch console and request the same interval through MCP. Compare event counts and raw entries, then expand the search to a related service. Use permissions appropriate to log analysis and check query limits and CloudWatch charges for the selected workflow.
Check the linked publisher documentation and your connected client's tool list. Names and availability can vary with the server version and account permissions.
| Tool | What it does |
|---|---|
| list_log_groups | List CloudWatch log groups with optional prefix and pagination. |
| search_logs | Search a selected log group over a bounded interval. |
| search_logs_multi | Search multiple log groups. |
| filter_log_events | Retrieve events matching a filter pattern. |
| summarize_log_activity | Summarize activity within a log group and interval. |
| find_error_patterns | Identify recurring error patterns in retrieved logs. |
| correlate_logs | Correlate matching events across selected log groups. |
{
"mcpServers": {
"cw-mcp-server": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/awslabs/Log-Analyzer-with-MCP",
"cw-mcp-server"
]
}
}
}Uses the normal AWS credential chain. Add the documented --profile and --region arguments when needed.
Python 3.12 or later, uv and AWS credentials with access to the selected CloudWatch Logs resources.
Give your coding agent the full DevTools toolbox: traces, network, console, heap snapshots and Lighthouse.
Workers, KV, R2 and D1 by conversation — Cloudflare's official remote servers for building and observability.
WaveGuard is a general-purpose anomaly detection API. Send it any data — server metrics, financial transactions, log files, sensor readings, time series — and get back anomaly scores, confidence…
A Model Context Protocol (MCP) server that provides seamless integration between any MCP client and the SpaceTraders API, enabling automation
MCP server that decodes minified production stack traces back to original TypeScript source files, lines, and columns using source maps.
Source profile for sonar-mcp-server. Its purpose and connection details need verification against the maintainer’s documentation.