azure-diagnostics
Debug Azure production issues on Azure using AppLens, Azure Monitor, resource health, and safe triage. WHEN: debug production issues, troubleshoot container apps, troubleshoot functions, troubleshoot AKS, kubectl cannot connect, kube-system/CoreDNS failures, pod pending, crashloop, node not ready, upgrade failures, analyze logs, KQL, insights, image pull failures, cold start issues, health probe failures, resource health, root cause of errors.
83
78%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./plugin/skills/azure-diagnostics/SKILL.mdQuality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong skill description that clearly defines its scope (Azure production debugging), lists specific tools and actions, and provides an extensive WHEN clause with natural trigger terms covering a wide range of failure scenarios. The description is well-structured, uses third person voice, and would be easily distinguishable from other skills in a large skill library.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and tools: AppLens, Azure Monitor, resource health, safe triage, and references specific scenarios like crashloop, pod pending, node not ready, upgrade failures, cold start issues, health probe failures, and KQL analysis. | 3 / 3 |
Completeness | Clearly answers both 'what' (debug Azure production issues using AppLens, Azure Monitor, resource health, safe triage) and 'when' with an explicit 'WHEN:' clause listing extensive trigger scenarios covering container apps, functions, AKS, and specific failure modes. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'debug production issues', 'troubleshoot container apps', 'kubectl cannot connect', 'CoreDNS failures', 'pod pending', 'crashloop', 'node not ready', 'image pull failures', 'cold start issues', 'health probe failures', 'KQL', 'analyze logs'. These are highly natural phrases a user would type. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche: Azure production debugging specifically for container apps, functions, and AKS. The combination of Azure-specific tools (AppLens, Azure Monitor) and Kubernetes-specific failure modes (CoreDNS, crashloop, pod pending) makes it very unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
57%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill has strong structural organization with excellent progressive disclosure and clear routing to service-specific guides. However, it suffers from some unnecessary verbosity (the mandatory compliance banner, generic diagnosis flow, trigger list) and the actionability could be improved with more precise MCP tool syntax and concrete error-recovery workflows. The diagnostic workflow lacks validation checkpoints expected for production incident triage.
Suggestions
Remove the 'AUTHORITATIVE GUIDANCE' banner and the 'Triggers' section — these waste tokens on information Claude can infer from context and frontmatter.
Replace the generic Quick Diagnosis Flow with a concrete decision tree: e.g., 'If HTTP 5xx → check resource health first → if healthy → query AppInsights for exceptions → if unhealthy → check activity log for platform events'.
Add explicit validation/feedback loops to the workflow: after attempting remediation, specify how to verify the fix worked (e.g., 'After restarting, re-query logs to confirm errors stopped').
Standardize MCP tool examples into a consistent, copy-paste-ready format or clarify the exact invocation syntax expected by the MCP tools.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The 'AUTHORITATIVE GUIDANCE — MANDATORY COMPLIANCE' banner is unnecessary filler. The 'Triggers' section largely restates what Claude can infer from context. The 'Quick Diagnosis Flow' is generic knowledge Claude already possesses. However, the tool invocation examples and CLI commands are appropriately concise. | 2 / 3 |
Actionability | CLI commands and MCP tool invocation patterns are concrete and useful, but the MCP examples use a pseudo-format rather than executable/copy-paste-ready syntax. The Quick Diagnosis Flow is vague ('What's failing?', 'What do logs show?') rather than providing specific actionable steps. Placeholder values like RESOURCE_ID and RG are appropriate but the overall guidance leans toward describing rather than fully instructing. | 2 / 3 |
Workflow Clarity | The Quick Diagnosis Flow provides a sequence but lacks validation checkpoints or feedback loops. There's no explicit 'if this fails, try that' error recovery guidance. The routing table is clear, but the actual diagnostic workflow doesn't include verification steps after remediation attempts, which is important for production debugging. | 2 / 3 |
Progressive Disclosure | Excellent use of progressive disclosure with a clear overview, routing table directing to service-specific guides, and well-signaled one-level-deep references (container-apps/, functions/, aks-troubleshooting/, kql-queries.md, azure-resource-graph.md). The content is appropriately split between the parent skill and child references. | 3 / 3 |
Total | 9 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- microsoft/github-copilot-for-azure
- Commit
a46a937
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.