26 tools with mixed quality. Naming is generally verb-first and clear (kubectl_*, install_*, start_*). Descriptions present but often generic (e.g., 'Get Kubernetes resources' lacks context on when to use vs kubectl_list). Schemas visible for all tools with proper JSON Schema structure, but many parameters lack descriptions or constraints. Critical gaps: no enum constraints on string parameters (searchMode, output, patchType have enums but most don't); parameter descriptions are sparse or missing (e.g., kubectl_search 'query' has no guidance on syntax); no output schemas documented; error handling minimal. kubectl_generic is a catch-all that undermines tool composition. Destructive tools (cleanup, kubectl_delete, uninstall_helm_chart) lack confirmation/dry-run patterns.

• Add output schemas to all tools. Document field names, types, and structure so agents can chain results. Example: kubectl_get should return {items: [{name, namespace, status, age}], total_count}.
• Enhance parameter descriptions with constraints and examples. Replace 'Output format' with 'Output format (json, yaml, wide, or name). Default: json'. Add regex patterns or enums for string parameters.
• Implement dry-run and confirmation patterns for destructive tools. kubectl_delete and cleanup should support dryRun=true and require explicit confirmation before executing.
• Remove or gate kubectl_generic. If needed, restrict to read-only commands or require explicit allowlist. Unrestricted arbitrary kubectl access breaks safety and tool composition.
• Add structured error responses with recovery guidance. Instead of raw errors, return {error: 'Resource not found', suggestion: 'Try kubectl_search to find available resources', retryable: false}.