Agents
| Agent | What it does | Tools |
|---|---|---|
| lab-author | Maintainer agent that writes and updates the self-paced lab pages in site/ so they match the lab template, the checks in lab/checks.yaml, and the files in the repo. Use to draft a new lab, fix a lab-feedback issue, or update pages after scripts, agents, or skills change. | read, edit, search, execute, todo |
| model-assessor | Reviews the Contoso Pharmacy semantic model against the team conventions and AI-readiness checklist, without changing it, and writes prioritized findings to out/assessment.md. Use before improving the model or to explain why AI answers are weak. | read, search, execute, edit, todo, powerbi-modeling-mcp/* |
| model-improver | Improves the Contoso Pharmacy semantic model for people and AI tools, including business names, descriptions, hidden keys, display folders, relationships, the date table, core measures, and Prep data for AI drafts, then rebinds the report and validates. Use after model-assessor, or when asked to fix a specific convention. | read, edit, search, execute, todo, powerbi-modeling-mcp/* |
| model-reviewer | Reviews semantic model and report changes on a branch or pull request like a senior Power BI developer, runs the local validation, and returns a review with blocking issues, suggestions, and questions. Use before opening or merging a pull request that touches fabric/. | read, search, execute |
| question-tester | Answers the five Contoso Pharmacy business questions using only the semantic model (metadata and DAX through Power BI Modeling MCP), saves the answers as evaluation/runs/<label>.json, and grades the run. Use to create a baseline or after run, or to test whether a model change improved answers. | read, search, execute, edit, todo, powerbi-modeling-mcp/* |
lab-author
Section titled “lab-author”Maintainer agent that writes and updates the self-paced lab pages in site/ so they match the lab template, the checks in lab/checks.yaml, and the files in the repo. Use to draft a new lab, fix a lab-feedback issue, or update pages after scripts, agents, or skills change.
- File: .github/agents/lab-author.agent.md
- Argument hint: Lab number or feedback issue, for example “Lab 07” or “#42”
- Tools:
read,edit,search,execute,todo
model-assessor
Section titled “model-assessor”Reviews the Contoso Pharmacy semantic model against the team conventions and AI-readiness checklist, without changing it, and writes prioritized findings to out/assessment.md. Use before improving the model or to explain why AI answers are weak.
- File: .github/agents/model-assessor.agent.md
- Argument hint: Optional focus, for example “names only” or “measures and time intelligence”
- Tools:
read,search,execute,edit,todo,powerbi-modeling-mcp/*
Handoffs (buttons that appear after the agent answers):
| Button | Goes to | Sends the prompt automatically |
|---|---|---|
| Fix these findings | model-improver | no, you review it first |
model-improver
Section titled “model-improver”Improves the Contoso Pharmacy semantic model for people and AI tools, including business names, descriptions, hidden keys, display folders, relationships, the date table, core measures, and Prep data for AI drafts, then rebinds the report and validates. Use after model-assessor, or when asked to fix a specific convention.
- File: .github/agents/model-improver.agent.md
- Argument hint: What to fix, for example “names and descriptions” or “the findings in out/assessment.md”
- Tools:
read,edit,search,execute,todo,powerbi-modeling-mcp/*
Handoffs (buttons that appear after the agent answers):
| Button | Goes to | Sends the prompt automatically |
|---|---|---|
| Re-test the business questions | question-tester | no, you review it first |
| Review the change | model-reviewer | no, you review it first |
model-reviewer
Section titled “model-reviewer”Reviews semantic model and report changes on a branch or pull request like a senior Power BI developer, runs the local validation, and returns a review with blocking issues, suggestions, and questions. Use before opening or merging a pull request that touches fabric/.
- File: .github/agents/model-reviewer.agent.md
- Argument hint: Base branch to compare with (default main)
- Tools:
read,search,execute
question-tester
Section titled “question-tester”Answers the five Contoso Pharmacy business questions using only the semantic model (metadata and DAX through Power BI Modeling MCP), saves the answers as evaluation/runs/<label>.json, and grades the run. Use to create a baseline or after run, or to test whether a model change improved answers.
- File: .github/agents/question-tester.agent.md
- Argument hint: Run label, for example local-baseline or local-ai-ready
- Tools:
read,search,execute,edit,todo,powerbi-modeling-mcp/*
Handoffs (buttons that appear after the agent answers):
| Button | Goes to | Sends the prompt automatically |
|---|---|---|
| Compare with the baseline | question-tester | no, you review it first |
| Review the change | model-reviewer | no, you review it first |