Agentic observability with Amazon OpenSearch Service MCP Apps
Machine Learning Blog
This article introduces Amazon OpenSearch Service MCP Apps, which extend the Model Context Protocol to deliver interactive observability visualizations alongside AI agent responses in IDE chat windows.
- MCP Apps render trace waterfalls, service maps, and log views directly in the IDE without browser tab-switching
- Dual response pattern combines text summaries with deterministic visualizations executed against actual data sources
- Local MCP server bridges IDE and OpenSearch UI while maintaining data security and credential control
- Supported tools include alert triage, log investigation, trace analysis, metrics, topology, and LLM observability
- Setup requires Node.js 22+, AWS credentials with ESHttp permissions, and configuration in Claude Desktop, VS Code, Cursor, or other IDEs
- Verification happens inline in a single conversation thread, reducing on-call investigation time from minutes to seconds
MCP Apps close the verification gap in agentic observability by keeping engineers in their IDE throughout investigation, verification, and resolution workflows.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
2026
2026
2026
2026
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.