Lumino MCP Server v0.9.2

[0.9.2] - 2026-01-24

Added

• Published to MCP Registry as io.github.geored/lumino
• Added PyPI badge to README
• Added mcp-name comment to README for registry validation

Changed

• Registry name changed to io.github.geored/lumino for publishing

[0.9.1] - 2026-01-24

Added

• MCP Registry support with server.json for registry submission
• [tool.mcp] section in pyproject.toml for registry integration
• Published to PyPI with full README documentation

Changed

• Updated package description to better reflect SRE observability capabilities
• Added keywords: sre, observability, pipelines

Fixed

• analyze_failed_pipeline - Handle deleted pods gracefully with fallback to TaskRun step info
• check_cluster_certificate_health - Fix duplicate namespace entries and respect user namespace filter
• check_resource_constraints - Add lowercase 'k' suffix support for count quotas (e.g., "2k" = 2000)
• detect_log_anomalies - Fix pattern key extraction showing "?i)" instead of category names
• ci_cd_performance_baselining - Filter out "unknown" task entries from Prometheus metrics

[0.9.0] - 2026-01-18

Added

Core Infrastructure

• Initial release of Lumino MCP Server with 37 MCP tools
• MCP (Model Context Protocol) integration for AI-powered Kubernetes operations
• Multi-cluster support with automatic context detection
• Prometheus integration for metrics queries
• Namespace caching with 1-day TTL for performance

Kubernetes Tools

• list_namespaces - List all namespaces in the cluster
• list_pods_in_namespace - List pods with status and placement info
• get_kubernetes_resource - Retrieve details about any Kubernetes resource
• search_resources_by_labels - Search resources across types and namespaces
• check_resource_constraints - Identify resource bottlenecks

Tekton Pipeline Tools

• list_pipelineruns - List PipelineRuns with status and timing
• list_taskruns - List TaskRuns, optionally filtered by PipelineRun
• find_pipeline - Find pipelines matching a pattern across namespaces
• get_pipelinerun_logs - Fetch logs from all pods in a PipelineRun
• get_tekton_pipeline_runs_status - Cluster-wide status summary
• list_recent_pipeline_runs - Recent PipelineRuns across all namespaces
• analyze_failed_pipeline - Root cause analysis for failed pipelines

Analysis & Diagnostics

• analyze_logs - Extract error patterns and insights from logs
• detect_anomalies - Statistical anomaly detection in PipelineRuns
• detect_log_anomalies - ML-powered log anomaly detection
• smart_get_namespace_events - Adaptive event analysis with auto-filtering
• progressive_event_analysis - Multi-level event correlation
• advanced_event_analytics - ML patterns with runbook suggestions

Log Analysis Tools

• smart_summarize_pod_logs - Adaptive pod log analysis
• stream_analyze_pod_logs - Streaming log analysis with pattern detection
• analyze_pod_logs_hybrid - Intelligent strategy selection for log analysis
• semantic_log_search - Natural language log search

Predictive & Forecasting

• predictive_log_analyzer - ML-based failure prediction
• resource_bottleneck_forecaster - Resource exhaustion forecasting
• what_if_scenario_simulator - Impact simulation for config changes

CI/CD Performance

• ci_cd_performance_baselining_tool - Performance baselines with statistical analysis
• pipeline_tracer - Trace operations through pipeline flows
• automated_triage_rca_report_generator - Automated root cause analysis reports

OpenShift Support

• get_machine_config_pool_status - MCP status and update monitoring
• get_openshift_cluster_operator_status - Cluster operator health checks
• get_etcd_logs - Retrieve etcd pod logs
• check_cluster_certificate_health - Certificate expiration scanning
• investigate_tls_certificate_issues - TLS issue investigation

Namespace Investigation

• conservative_namespace_overview - Quick namespace health check
• adaptive_namespace_investigation - Progressive namespace analysis
• live_system_topology_mapper - Real-time dependency graph generation

Prometheus Integration

• prometheus_query - Execute PromQL queries with automatic auth

Performance Optimizations

• find_pipeline - Cluster-wide queries with API limits and optional TaskRun fetching
• ci_cd_performance_baselining_tool - Parallelized Prometheus queries
• get_tekton_pipeline_runs_status - Configurable limits to prevent timeouts on large clusters
• pipeline_tracer - Parallelization and namespace targeting
• predictive_log_analyzer - Namespace targeting and improved log collection

Container Image

• Available at quay.io/geored/lumino-mcp-server
• Multi-architecture support (amd64, arm64)

Container Image

docker pull quay.io/geored/lumino-mcp-server:0.9.2

Installation

See README.md for installation and configuration instructions.