infrastructure incident triage

    1

    Diagnose infrastructure outages across Linux, K8s, and networking gear using read-only evidence-based triage.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+22 more

    infrastructure incident triage

    Example session with this skill installed

    The checkout service is returning 500 errors. It started 10 minutes ago. Triage the infrastructure to find the cause.

    • Read your context and instructions
    • Compiled the infrastructure incident triage

    I have analyzed the checkout service incident. The triage covered the Kubernetes cluster, underlying Proxmox hosts, and the Cisco core switch.

    Incident Triage Report: Checkout Service 500 Errors

    HypothesisConfidenceSupporting EvidenceContradicting EvidenceBest Next Check
    Node Disk PressureHighk8s_get_node_status shows node-04 DiskPressure; proxmox_get_storage shows 98% utilization on ssd-pool-01.host_get_memory shows 40% free; no CPU spikes in cisco_get_stats.Check ssd-pool-01 for orphan snapshots.
    Pod Scheduling FailureMediumk8s_get_events shows 12 FailedScheduling events for checkout-api pods in the last 8 mins.None.Verify node affinity rules.
    Database LatencyLowgrafana_get_alert shows a spike in DB query time (250ms), but it trails the 500 errors by 2 mins.docker_get_stats for the DB container shows stable I/O.Check DB slow query logs.

    Evidence Timeline

    1. 14:02:15 - cisco_get_port_stats: No errors on uplink. Network layer healthy.
    2. 14:04:30 - proxmox_get_storage: ssd-pool-01 exceeds 95% threshold.
    3. 14:05:10 - k8s_get_node_status: node-04 transitions to Ready: False, DiskPressure: True.
    4. 14:06:00 - k8s_get_events: checkout-api pods evicted from node-04.

    Coverage Gaps

    • Could not inspect redis-cache as the specific Docker adapter for that instance was unreachable.

    Next steps

    • Review Proxmox storage ssd-pool-01 to identify and delete unnecessary large files or old snapshots.
    • Increase the disk quota for node-04 or add a new worker node to the cluster.
    • Adjust K8s eviction thresholds to provide more buffer for the checkout service.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Infrastructure issues often lead to wasted hours chasing red herrings across disparate systems. This skill formalizes the triage process by interfacing directly with your environment through the Infrastructure MCP server. It systematically investigates Linux hosts, Kubernetes clusters, Docker containers, and networking hardware to find the source of degradation without manual SSH hopping or log chasing.

    What it does

    • Layered diagnostics scans network, host, container, and platform layers to map the incident scope.
    • Evidence-based reporting correlates timestamps across Proxmox, K8s, and Cisco devices to build a timeline.
    • Hypothesis ranking assigns confidence levels to potential root causes based on tool-verified signals.
    • Read-only safety ensures no state changes or mutations occur during the investigative phase.
    • Gap identification explicitly flags systems or layers that could not be inspected due to missing adapters.

    How it works

    1. Frame the incident by defining symptoms, affected services, and the start time of the anomaly.
    2. Scope the targets using the list_servers tool to identify relevant hosts, clusters, and instances.
    3. Execute broad-to-narrow collection of alerts, status metrics, host health, and targeted logs.
    4. Correlate events by ordering signals chronologically to find the earliest point of failure.
    5. Generate the report with ranked hypotheses and clear evidence citations for each claim.

    Frameworks & tools

    Requires a running Infrastructure MCP Server. Compatible with Linux, Docker, Kubernetes, Grafana, Proxmox, and Cisco networking environments.

    Why this beats prompting it yourself

    Standard prompts often jump to conclusions or suggest invasive commands that can worsen an outage. This skill enforces strict evidence discipline, requiring two independent signals for root cause claims and maintaining a strictly read-only posture to protect production stability.

    Use cases

    • Diagnose sudden latency spikes in a Kubernetes microservice.
    • Investigate unexplained reboots or resource exhaustion on Linux hosts.
    • Trace network connectivity issues across Cisco switches and Proxmox VMs.
    • Analyze firing Grafana alerts to find the underlying infrastructure trigger.

    Known limitations

    Requires an active Infrastructure MCP server connection. Limited to read-only diagnostics; cannot perform remediation or state changes. Tail and limit values are capped to ensure performance.

    How to use

    Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill passes our trust checks below.

    Security scanned

    Passed our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scanned

    Verified clean 5 days ago

    • Passed all security checks, Safe to install

    Requires

    Infrastructure MCP

    Listed5 days ago

    What's inside

    Frequently Asked Questions