This project provides an intelligent error monitoring agent that leverages Airweave to automatically enrich raw error data with context from code repositories, ticketing systems, and communication platforms. It targets engineers and power users seeking to transform noisy alerts into actionable insights by semantically clustering similar errors, identifying root causes, and determining appropriate severity and alerting actions, thereby reducing alert fatigue and improving incident response times.

How It Works

The agent processes raw errors through a multi-stage pipeline: semantic clustering to group similar issues, context search via Airweave to find related code, tickets, and Slack discussions, and finally, analysis to determine severity and status. This approach is advantageous as it moves beyond simple error aggregation to provide rich, contextual understanding, enabling smarter deduplication and suppression logic that prioritizes genuinely novel or critical issues.

Quick Start & Requirements

To run the interactive demo: clone the repository, copy .env.example to .env, optionally add an OPENAI_API_KEY or ANTHROPIC_API_KEY for enhanced clustering, navigate to the backend directory, install dependencies (pip install -r requirements.txt), and start the backend server (uvicorn main:app --reload --port 8000). Then, in the frontend directory, run npm install && npm run dev. Access the demo at http://localhost:3000. Production setup requires configuring DATA_SOURCE (Sentry, Azure, or custom), AIRWEAVE_API_KEY, and optionally LINEAR_API_KEY and SLACK_BOT_TOKEN for integrations.

Highlighted Details

• Multi-Stage Clustering: Errors are initially grouped by type (e.g., HTTP status codes, exception classes), then merged, with an optional LLM stage for deeper semantic similarity analysis.
• Smart Suppression Logic: Alerts are intelligently suppressed for ongoing issues with open tickets, muted error signatures, or recent notifications, while prioritizing NEW or REGRESSION errors and S1/S2 severity issues.
• Context Enrichment: Airweave is used to automatically search connected GitHub repositories, Linear tickets, and Slack threads for context relevant to each error cluster, aiding immediate understanding.
• Production Integrations: Supports direct integration with Sentry or Azure Log Analytics for error ingestion, and enables automated ticket creation in Linear and notifications in Slack.

Maintenance & Community

The project is based on an internal agent ("Donke") that handles significant Airweave query volume monthly. No specific community channels (Discord, Slack) or explicit contributor information are detailed in the README.

Licensing & Compatibility

The project is released under the MIT License, which is permissive for commercial use and integration into closed-source applications.

Limitations & Caveats

The primary focus is on the demonstration and setup of the error monitoring pipeline. Production deployment requires significant configuration and integration with external services like Airweave, error tracking platforms (Sentry, Azure), and communication tools (Linear, Slack). LLM-based clustering is optional but recommended for optimal performance. The README does not explicitly state an alpha or beta status, but the emphasis on a demo suggests it may be early-stage.