BugZooka

Overview

BugZooka operates as a Slack monitoring and conversational bot designed to parse and categorize log errors using configurable prompts and OpenAI-compatible LLM endpoints. It supports both a polling mode for routine channel failure assessments and a real-time Socket Mode for chat interactions. Through integration with external tools like the Orion MCP server, it can run advanced multi-step analysis on GitHub pull requests, check nightly test regressions, and optionally incorporate RAG-based context from local vector stores to deliver rich infrastructure insights.

Capabilities

• ▸Slack channel log monitoring and automated failure summarization
• ▸LLM-powered log analysis using custom or default prompts
• ▸Real-time event processing and command handling through Slack Socket Mode
• ▸GitHub pull request performance analysis via Orion-MCP integration
• ▸Nightly test regression detection and comparison
• ▸Optional RAG-augmented insights using local vector stores

Best for

Analyzing and categorizing CI job failure logs directly within Slack channels, Generating performance summaries and trend comparisons for Kubernetes and OpenShift across specified lookback windows, Inspecting nightlies for test regressions and comparing versions using Model Context Protocol (MCP) tool calls, Interactively evaluating pull request performance impacts via real-time Slack bot mentions

Works with

Security considerations

Users should avoid inputting personal or sensitive corporate information into LLM prompt inputs. All necessary authorization credentials, API keys, and Slack tokens must be securely stored in a local .env configuration file.

Categories