Creates Azure Architecture Decision Records with WAF mapping, alternatives, and consequences. USE FOR: ADR creation, architecture decisions, trade-off analysis, WAF pillar justification. DO NOT USE FOR: Bicep/Terraform code generation, diagram creation, cost estimates.
77
71%
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 ./.github/skills/azure-adr/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 an excellent skill description that clearly defines its scope with specific actions, strong trigger terms, and explicit inclusion/exclusion boundaries. The USE FOR and DO NOT USE clauses make it easy for Claude to select this skill accurately and avoid conflicts with related Azure infrastructure skills. The description is concise yet comprehensive.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: creating ADRs, WAF mapping, alternatives analysis, and consequences documentation. Also specifies the Azure context clearly. | 3 / 3 |
Completeness | Clearly answers both 'what' (creates Azure ADRs with WAF mapping, alternatives, consequences) and 'when' (explicit USE FOR clause with trigger scenarios). The DO NOT USE clause adds further clarity on boundaries. | 3 / 3 |
Trigger Term Quality | Includes strong natural trigger terms users would say: 'ADR', 'architecture decisions', 'trade-off analysis', 'WAF pillar', 'Architecture Decision Records'. The DO NOT USE section also helps prevent false triggers on related but distinct tasks like 'Bicep', 'Terraform', 'diagram'. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche (Azure ADRs with WAF mapping). The explicit DO NOT USE clause for Bicep/Terraform code generation, diagrams, and cost estimates actively prevents conflicts with adjacent infrastructure skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
42%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill has excellent progressive disclosure with well-signaled references and a clear reference index. However, it suffers from significant verbosity with redundant sections (workflow integration appears twice, trigger phrases and 'does not do' sections are unnecessary), and the core actionable content (the actual ADR template) is entirely deferred to a reference file without any inline example of expected output. The workflow is present but lacks feedback loops for error recovery.
Suggestions
Remove redundant sections: consolidate the two 'Workflow Integration' sections into one, and remove the 'When to Use This Skill' trigger table and 'What This Skill Does NOT Do' section (already covered by the YAML description).
Include a brief concrete example of a completed ADR output (even abbreviated) so Claude has an inline reference without needing to load the template file.
Add a feedback loop to the Generation Workflow, e.g., 'After generating, run through the Quality Checklist; if any item fails, revise the ADR before saving.'
Remove the 'Example Prompts' section — these describe how users prompt Claude, not how Claude should execute the skill.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Significant verbosity and redundancy throughout. The 'When to Use This Skill' trigger table, 'Common ADR Topics' table, and 'What This Skill Does NOT Do' section explain things Claude can infer. The 'Workflow Integration' section appears twice with overlapping content. Multiple tables restate the same design vs as-built distinction. The 'Example Prompts' section describes how to prompt Claude, which is unnecessary in a skill file. | 1 / 3 |
Actionability | The skill provides a clear generation workflow, naming conventions, and a quality checklist, which are concrete. However, the actual ADR template content is deferred to a reference file, meaning the core executable guidance is not present in the skill itself. No concrete example of a completed ADR output is shown, only example prompts. | 2 / 3 |
Workflow Clarity | The 'Generation Workflow' section lists 6 steps in a clear sequence, and the quality checklist serves as a validation checkpoint. However, there are no explicit feedback loops (e.g., what to do if the checklist fails, how to handle missing information). The workflow lacks validation-then-fix-then-retry patterns. | 2 / 3 |
Progressive Disclosure | References are well-structured with a clear 'Reference Index' table at the bottom, explicitly stating when to load each file. The instruction 'Load these on demand — do NOT read all at once' is excellent. References are one level deep and clearly signaled throughout the document. | 3 / 3 |
Total | 8 / 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.
ec7b8ff
Table of Contents
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.