SuperAPI AI Directory

digma-mcp-server

A code observability MCP enabling dynamic code analysis based on OTEL/APM data to assist in code reviews, issues identification and fix, highlighting risky code etc.

Open website

About this tool

Digma Code Observability MCP Server is a Model Context Protocol (MCP) server implementation designed to enable agents to leverage Digma to access observability insights, conduct code observability and dynamic code analysis.

Purpose:

  • Observability-assisted code review: Inspect issues found by pre-production observability in PR branches.
  • Discover code inefficiencies through dynamic code analysis: Identify code/query issues that slow down the application.
  • Leverage distributed tracing for code runtime usage data: Inspect breaking changes or generate related tests.

Key capabilities (based on example prompts):

  • Help review code changes in a branch and view related runtime issues.
  • Identify the top three issues affecting application performance the most for improvement.
  • Based on runtime data, understand which other services and code will be affected by modifying a certain function.
  • Check whether the code has new issues in the staging environment.
  • Identify the database queries with the greatest impact on application performance.

How to use:

  • Configure an MCP client (such as Claude, Cursor, etc.) to include Digma MCP. A Digma deployment includes an MCP SSE server. You can configure it in the client using its URL, or use an MCP tool such as SuperGateway to run it as a command tool.
  • The MCP URL path consists of the Digma API Key: https://<DIGMA_API_URL>/mcp/<DIGMA_API_TOKEN>>/sse
  • If the client supports SSE servers, the corresponding syntax can be used; if used as a command tool, it can be bridged to the URL via SuperGateway.
  • The agent autonomously chooses when to use the data provided by Digma, but some clients allow rules and policies to be set to establish a more structured workflow. Example rule files can be added to Cursor's .cursor/rules directory.

Applicable scenarios:

  • Development teams that need to combine runtime observability data for code review and issue localization.
  • Scenarios where distributed tracing data is expected to be used to optimize application performance and database queries.
  • Scenarios where the impact of code changes is checked in a pre-production or staging environment.

Limitations:

  • Requires Digma early access, and you need to visit the MCP page to register.
  • Depends on the Digma deployment and API Key.
  • Specific feature details and availability maybe limited, and the official offering shall prevail.

License: MIT License, see the LICENSE file for details.