Prompt file imported from nileshgule1981/MS-Cert-Buddy-Agent (.github/prompts/ai102-practice-lab.prompt.md). Fill in {{skillArea}}, {{objective}}, {{toolPreference}}, {{timebox}} before use. Copyright stays with the author.

AI-102 Practice Lab

Generate ONE short, self-validating AI-102 practice lab.

Use this skill

You must follow the workspace skill ai102-lab-creator for lab structure, guardrails, workflow, output format, and delivery rules (full lab in a single message).

Inputs (from chat)

• Skill area: {{skillArea}}
• Objective: {{objective}}
• Tool preference: {{toolPreference}}
• Timebox: {{timebox}}

Grounding and validation rules

1. Ground the lab outcome in Microsoft Learn using Context7 tools (which index Learn documentation) and Copilot web search.
2. Validate every lab step with Azure MCP tools: resource providers exist, properties and flags are real, validation queries detect success or failure, and cleanup commands are correct.
3. If the lab includes CLI or PowerShell syntax, confirm command names, flags, and module versions with Context7 tools.
4. Provide Microsoft Learn URLs in the References section of the lab output.

Terminology

Always use current Microsoft product names. Never use retired names (for example, "Azure AD" must be "Microsoft Entra ID"). See the full rename table in .github/copilot-instructions.md. If a lab touches identity or governance, double-check every product name against current terminology.

Tool preference default

This prompt is designed for Azure CLI practice labs. If the user does not specify a tool preference, default to Azure CLI as the primary instruction path. When Azure CLI is the path, steps must use exact az commands with correct flags and parameters.

Output format

Use the YAML output format defined in the ai102-lab-creator skill exactly. The output must include: title, objective, skill_area, estimated_time, prerequisites, starting_state, tasks (each with steps and validation), troubleshooting, cleanup (with validation), and references.

Style rules

• No contractions.
• No ambiguous "click around until you find it" steps. Every portal step must name the exact blade, menu, or button label.
• Use Microsoft instruction formatting conventions for UI labels (bold for clickable elements, exact casing).
• Distractors in troubleshooting must describe real failure modes, not invented ones.
• All Azure product names must use current terminology (no retired names).
• Cleanup is mandatory. Every lab must end with steps that remove all created resources and a validation gate that confirms removal.
• Keep the lab within AI-102 scope (AI solution design and implementation, not general Azure administration).
• Prefer lowest-cost resources; include a cost warning if the lab uses resources that incur charges beyond the free tier.