Query Loki logs, Prometheus/Mimir metrics, and Tempo traces to investigate production issues.
npx skills add https://github.com/pokgak/agent-skills --skill lgtm-pokgak
Please help me install this Agent Skill. Skill: lgtm Source: https://github.com/pokgak/agent-skills/tree/main/skills/lgtm Command: npx skills add https://github.com/pokgak/agent-skills --skill lgtm-pokgak
Quickly investigate production incidents by correlating logs, metrics, and distributed traces instead of manually guessing where the failure started.
Use when you need to answer questions like: Why are errors increasing, what changed in latency, where is the request spending time, and which services are impacted?
Ask the agent to: Investigate why the checkout app is seeing elevated error rates in production over the last hour using logs, metrics, and traces.
High-intent search queries and answers about installing and using this skill.
Debug production incidents by querying Loki logs, Prometheus metrics, and Tempo traces to correlate observability signals across services. The tool orchestrates discovery of labels and services before running targeted queries to identify root causes for errors and latency regressions.
Investigate elevated error rates by orchestrating a discovery-and-investigation workflow that searches Loki logs, analyzes Prometheus metrics, and inspects Tempo traces. It extracts actionable signals like error counts, trends, and top offenders into concise summaries instead of raw JSON.
Yes, you need a valid lgtm configuration at ~/.config/lgtm/config.yaml or Grafana Cloud lgtm discover to query Loki logs and Prometheus metrics. This configuration provides the necessary endpoints and authentication for the observability querying workflow.
Find where request latency is spending time by inspecting Tempo distributed traces using TraceQL search patterns across time ranges and namespaces. The tool distills JSON trace outputs into concise summaries highlighting trace IDs and latency bottlenecks for faster decision-making.
Yes, you can analyze service health regressions across specific time ranges and namespaces by querying Prometheus metrics and Loki logs. The tool applies a context-efficient orchestration pattern to gather valid label values before running targeted health-oriented queries.
Raw JSON outputs from PromQL queries are hard to use because they obscure the signal. This tool distills JSON outputs from logs, metrics, and traces into concise summaries, extracting counts, trends, top offenders, and trace IDs for faster incident debugging and decision-making.
cosmixConfigures Grafana-based observability dashboards with the LGTM stack, including RBAC and alerting for logs, traces, and metrics.
BlackkadderFetch aggregated trace metrics from MLflow tracking servers.
grafanaDiagnoses application issues using Grafana metrics, logs, and traces.
xobotyiExplain logs, metrics, and traces for system observability.
