Logs Sieve MCP server

Logs Sieve is a small helper that connects your AI assistant to the place where your app keeps its logs. Instead of clicking around a log dashboard and guessing time windows, you just describe what you are looking for. It is handy for anyone who gets asked "why did this break?" and has to dig through log tools.

What this MCP server does

You ask your AI something like "show me payment errors from the last hour". The AI passes that request to Logs Sieve, which knows how to talk to your chosen log provider. Logs Sieve picks the right source, narrows the time window, and pulls back the matching entries. The AI then reads them and gives you a short answer, a list of entries, or a summary of what keeps going wrong. It can also follow one request across several services if you give it a trace id.

Click to zoom

What you can do with it

• Search logs by keyword, time window and scope
• Summarize what happened in a log window in a few lines
• Trace one request across services using a trace id
• List the log sources available in your project
• Spot recurring failure patterns without reading every line
• Ask follow-up questions about the same incident in the same chat

Try asking your AI

• “Show me all payment failed errors in my project from the last hour”
• “Summarize what went wrong in the checkout service over the last 24 hours”
• “Trace request id abc-123 across all services and tell me where it slowed down”
• “What log sources do I have available in this project?”

What it gives back to you

You get back plain answers in the chat: lists of matching log lines, short summaries of what happened, or a step by step trace of one request. Numbers like counts and time windows are included when they help. Everything stays read only, so nothing in your log system is changed.

Before you start

Who it's for

Engineers, support staff and on call people who need to look at logs during incidents but do not want to learn each log tool's query language.