• QUERY
  • FILTER
  • ANALYSIS
  • HUMAN

How to debug application errors with your AI

Connect your AI to Azure Log Analytics Workspace once. From there, you just ask questions to find what went wrong in your logs.

Ask AI about this page

Short answer

How can I use my AI to find error codes in Azure?

Your AI uses the query_logs tool to run KQL queries against your log tables. It pulls the specific failure codes and error messages you need to see. Here is how it handles the investigation.

The investigation outcomes.

Where your error data lands.

The AI does the searching so you can focus on the fix.

QUERY

Instant KQL generation

The AI turns your plain English into valid KQL. You get the exact syntax needed to hit your specific log tables.

FILTER

Targeted error isolation

The AI filters through massive datasets to find specific failure codes. You get a clean list of the errors that actually matter.

ANALYSIS

Pattern recognition

The AI looks at the returned logs to find trends. It spots if a specific error is spiking or tied to a certain resource.

HUMAN

Manual code fixes

The AI identifies the error, but you have to fix the code. You use the filtered results to decide how to patch the app.

The workflow

What your AI does when the logs arrive.

The AI acts as the interface between your questions and the Azure data.

  1. Translate intent

    Your AI takes your description of a problem and converts it into a structured query.

    query_logs

  2. Execute search

    The AI runs the generated KQL against your Azure Log Analytics Workspace to pull the raw data.

    query_logs

  3. Parse results

    It scans the returned rows to find the specific error messages or status codes you asked for.

    query_logs

  4. Summarize findings

    The AI presents the relevant error logs in a readable format, highlighting the most frequent failures.

    query_logs

Try it

Copy these to start.

Paste these into your chat to begin an investigation.

Starting points

Swap the table names and error codes in these prompts for your own environment.

Azure Log Analytics Workspace Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_h8mzxcilFYlijQMfr5u3GIGk82xs7W4r1G5Um6EK/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Azure Log Analytics Workspace capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "azure-log-analytics-workspace-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_h8mzxcilFYlijQMfr5u3GIGk82xs7W4r1G5Um6EK/mcp"
    }
  }
}
Copy into chat04

  • Run a KQL query to find all 'Error' level entries in the AppServiceConsoleLogs table from the last hour.

  • Search the MyCustomLogTable for any occurrences of error code 500 in the last 24 hours.

  • Find the most frequent exception messages in my web application logs for today.

  • Show me all logs in the SecurityEvent table that contain the word 'failed' from the last 30 minutes.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Azure Log Analytics Workspace once, then ask.

One link connects your AI to your workspace. Your login stays encrypted on our side, and you can start querying immediately.

Connect Azure Log Analytics Workspace to your AI

FAQ

How the debugging task behaves

  • 01

    Can the AI change my log settings?

    No. The AI can only read and query existing logs.

  • 02

    Does the AI delete any log data?

    No. The AI has no capability to delete or modify any data in your workspace.

  • 03

    Can I use this to create new logs?

    No. This connector is strictly for querying and retrieving data that already exists.

  • 04

    Do I need to write KQL myself?

    No. You can describe what you are looking for in plain English, and the AI will write the KQL for you.

  • 05

    Can the AI fix the application errors?

    No. The AI identifies the errors, but you must perform the actual code changes.

  • More questions about Azure Log Analytics Workspace? The Connector page answers them. See everything the Azure Log Analytics Workspace Connector can do

Connect Azure Log Analytics Workspace to Claude, Cursor, ChatGPT & more