Projects with this topic
-
A Go-based Open Source Universal Memory server that exposes a Model Context Protocol (MCP) interface for LLM clients. It provides semantic document search (for user documents), agent memory management, and a knowledge graph - backed by PgVector or ChromaDB - enabling AI agents to store, retrieve, and interconnect knowledge across sessions.
Updated -
Agentic EKS Clinician
Autonomous diagnostics for Amazon EKS. Agentic EKS Clinician is a Python-based operational analysis tool designed to detect, diagnose, and recommend fixes for Kubernetes and AWS infrastructure issues in EKS clusters. It moves beyond dashboards by performing structured analysis of cluster state and producing executive-ready HTML reports with evidence, diagnosis, and recommended actions.
North Star Detect → Diagnose → Recommend → Verify The goal is not observability. The goal is operational clarity and guided remediation.
What It Does (MVP) Agentic EKS Clinician analyzes: Kubernetes Pods, Nodes, and Events Scheduling failures Resource pressure and headroom risk OOMKilled containers AWS node group signals (via APIs)
It generates: Self-contained HTML reports JSON findings (machine-readable) Structured remediation guidance Current Diagnostic Packs EKS-RES-001 – Resource/OOMKilled detection EKS-NODE-001 – Node headroom risk EKS-SCHED-001 – Pending >2m scheduling analysis
Each pack: Identifies symptoms Explains likely root cause Provides recommended actions Includes verification commands
UpdatedUpdated