Skip to content

commands.md

Latest commit

 

History

History
262 lines (205 loc) · 7.83 KB

commands.md

File metadata and controls

262 lines (205 loc) · 7.83 KB

KDM CLI Command Reference

This document provides a comprehensive command reference for the KDM (Kubernetes Docker Monitor) CLI, detailing every command group, subcommand, option flag, and usage example.

📌 Global Options

These options apply when invoking the main kdm executable:

Flag Description
-v, --version Output the current version of KDM CLI
-h, --help Display general help and command listings

1. kdm show

Retrieve and display workload resources or cluster metadata in clean, formatted tables.

Subcommands:

  • kdm show runners Displays a combined table of all active Kubernetes pods and Docker containers.
  • kdm show pods Displays only Kubernetes pods.
  • kdm show containers Displays only Docker containers.
  • kdm show nodes Displays details of Kubernetes cluster nodes.
  • kdm show minikube Checks and prints local Minikube profile and connection status.

Examples:

kdm show runners
kdm show pods
kdm show minikube

2. kdm health

Checks the health of workloads and outputs color-coded logs and details.

Options:

  • -w, --watch: Run in watch mode to continuously refresh the terminal output.
  • -i, --interval <seconds>: Refresh interval in seconds (default: 5).

Subcommands:

  • kdm health all Unified health check of both containers and pods.
  • kdm health pods Health checks for pods only.
  • kdm health containers Health checks for Docker containers only.

Examples:

kdm health all
kdm health pods -w           # Continuous watch mode (5s refresh)
kdm health containers -w -i 2 # Continuous watch mode (2s refresh)

3. kdm watch

Launches the interactive real-time terminal user interface (TUI) dashboard. Refreshes status and metrics automatically (Ctrl+C to exit).

Example:

kdm watch

4. kdm logs <name>

Retrieves logs from a workload. KDM attempts a Docker container prefix match first, falling back to Kubernetes pods if no matching container is found.

Parameters:

  • <name>: The container name/ID prefix or the Kubernetes pod name.

Example:

kdm logs auth-api

5. kdm analyze

Scans Kubernetes resources for configuration errors and operational problems. Supports AI-powered troubleshooting explanations.

Options:

  • -n, --namespace <namespace>: Limit check to a specific Kubernetes namespace.
  • -L, --selector <selector>: Limit check to resources matching a label selector.
  • -f, --filter <filter>: Run a specific analyzer only (e.g. Pod, Ingress, Deployment). Can be specified multiple times.
  • -o, --output <format>: Output format choice: text (default) or json.
  • -m, --max-concurrency <num>: Max concurrency count for analyzers (default: 10).
  • -s, --with-stat: Print execution diagnostics statistics.
  • --with-doc: Retrieve Kubernetes documentation lookups for detected problems.
  • -e, --explain: Request AI-powered diagnosis explanations.
  • -b, --backend <backend>: Force a specific AI backend provider to query.
  • -l, --language <lang>: Request AI response in a target language (default: english).
  • -a, --anonymize: Mask resource names and identifiers in prompt payload to protect privacy.
  • -c, --no-cache: Skip looking up or saving to the local AI cache.
  • --kubeconfig <path>: Specify an alternative kubeconfig file path.
  • --kubecontext <context>: Specify target Kubernetes context.

Examples:

kdm analyze
kdm analyze -n default --explain
kdm analyze --explain --backend ollama --language spanish
kdm analyze -o json --anonymize

6. kdm auth

Manage credentials, models, and endpoints for AI backend providers. KDM supports 10+ AI providers (OpenAI, Anthropic, Gemini, Vertex AI, WatsonX, OCI GenAI, Cohere, Bedrock, Groq, HuggingFace, etc.).

Subcommands:

  • kdm auth add Register credentials for a provider.

    • Options:
      • -b, --backend <backend>: Target backend name (e.g. openai, ollama, anthropic, google-gemini, google-vertex, etc.).
      • -m, --model <model>: Target model name (e.g. gpt-4o, claude-3-5-sonnet-latest).
      • -p, --password <password>: API Key or password credential.
      • -u, --baseurl <baseurl>: Custom API base URL.
      • -t, --temperature <value>: Model generation temperature.
      • --topp <value>: Top-P sampling threshold.
      • --topk <value>: Top-K sampling threshold.
      • --maxtokens <value>: Max tokens limit.
      • --custom-header <Header=Value>: Add custom HTTP request header. Can be specified multiple times.

  • kdm auth list Lists all configured providers and their parameters (API keys are safely masked).

  • kdm auth default <backend> Sets the active default AI provider.

  • kdm auth remove <backend> Removes configured settings for a provider.

  • kdm auth update <backend> Updates settings on an already registered provider (supports the same options as add).

Examples:

kdm auth add -b openai -m gpt-4o -p sk-...
kdm auth add -b ollama -m llama3.1
kdm auth default ollama
kdm auth update ollama -t 0.2
kdm auth list

7. kdm cache

Manage locally cached AI explanations to save tokens and avoid duplicate network requests.

Subcommands:

  • kdm cache list Lists all cached explanation keys.
  • kdm cache get <key> Retrieves and prints cached markdown explanation for a key.
  • kdm cache remove <key> Deletes a specific cache entry.
  • kdm cache purge Deletes all locally cached explanations.

Examples:

kdm cache list
kdm cache purge

8. kdm filters

Configure default active analyzers to filter what kdm analyze checks.

Subcommands:

  • kdm filters list Lists all active default analyzers and available inactive ones.
  • kdm filters add <name> Adds an analyzer (e.g. Ingress, CronJob) to default active check list.
  • kdm filters remove <name> Removes an analyzer from the active list.

Examples:

kdm filters list
kdm filters add Ingress

9. kdm custom-analyzer

Register custom shell commands or HTTP webhooks to analyze arbitrary custom resources (CRDs).

Subcommands:

  • kdm custom-analyzer add <name> Register a new analyzer. Requires either --command or --url.

    • Options:
      • --command <cmd>: External CLI command to execute.
      • --url <url>: HTTP endpoint URL to call.

  • kdm custom-analyzer list Lists all registered custom analyzers.

  • kdm custom-analyzer remove <name> Deletes a registered custom analyzer.

Examples:

kdm custom-analyzer add keda --command "kubectl get scaledobjects -A -o json"
kdm custom-analyzer list
kdm custom-analyzer remove keda

10. kdm serve

Starts the KDM server daemon in either REST API mode or JSON-RPC Model Context Protocol (MCP) mode.

Options:

  • -p, --port <port>: HTTP server port (default: 8080).
  • --metrics-port <port>: Metrics server port.
  • -b, --backend <backend>: Default active AI backend provider.
  • -f, --filter <filter>: Default analyzer filter limit.
  • --http: Force HTTP mode (default).
  • --mcp: Start in MCP mode (JSON-RPC stdio).

Examples:

kdm serve --port 8080 # Starts HTTP REST API on http://localhost:8080
kdm serve --mcp       # Starts JSON-RPC MCP server over stdio

11. kdm config

Interactively set up alerting notifications or directly update config properties.

Subcommands:

  • kdm config setup Launches interactive CLI wizard to configure alerting notifications (Discord Webhook, Email SMTP settings).
  • kdm config list Lists all general configurations.
  • kdm config set <key> <value> Manually sets a config parameter.
  • kdm config clear Resets all configurations to default.

Examples:

kdm config setup
kdm config set alert_cooldown 600
kdm config list