Your First Log

After completing the Quick Start, verify that logs are flowing. From there, you can connect MCP so your AI agent can inspect the same production context.

Check the dashboard

1. Open your project at auralog.ai/projects.
2. Navigate to Logs in the sidebar.
3. You should see your logs appearing within a few seconds.

If no logs appear after 30 seconds, check that:

• Your ingest key is correct (starts with aura_).
• Your app is running and the init() call is executing.
• There are no network errors in the browser console or Node.js output.

Connect MCP

Use the MCP/API read key from the new-project startup modal, or create another read key under Settings → API & MCP keys. Then follow MCP Setup to connect Codex, Claude, Cursor, or another MCP client.

Once connected, ask something like:

• “Show me the most recent logs in production.”
• “Search for setup-test and summarize what happened.”
• “Find recent errors and group them by message.”

Optional: trigger an analysis

If you added an Anthropic or OpenAI key, auralogs can analyze errors automatically. To see that in action:

auralog.error("Test error for verification", {
  source: "setup-test",
});

After sending an error:

1. Go to Analyses in your project sidebar.
2. Within a minute or two, your AI provider will generate an analysis of the error.
3. The analysis includes severity, root cause assessment, and recommended actions.

If the analysis doesn’t appear, check Settings → AI providers — analyses are skipped when no provider key is configured or when the provider has rejected the key. The dashboard surfaces this with a banner at the top of the project.

What’s next

• MCP Setup — connect your AI agent to your logs.
• Configuration — customize capture behavior, flush intervals, and more.
• Error Capture — learn how automatic error capture works.
• AI Analysis — optionally choose between Anthropic and OpenAI, and understand what the model looks for.