Reduce early-workshop attrition at Step 7/14b and strengthen Step 5 active learning

[Copilot]

The simulation report shows concentrated dropout in Part 1, led by Copilot access friction at Step 7 and missing workflow-editing guidance at 14b, plus a weak active-learning signal in the Step 5 intro cluster. This PR tightens those specific instructional gaps without lowering practice depth.

• Step 7d: front-load Copilot access failure handling

  • Added a pre-flight decision tree before billing selection to route learners through account/org entitlement checks early.
  • Added a billing quick-reference table to clarify path selection and required configuration state (copilot-requests: write vs COPILOT_GITHUB_TOKEN).

• Step 14b: explicit agentic editing path for PR reviewer workflow

  • Added direct /agentic-workflows usage guidance when learners modify the workflow brief.
  • Added a trigger recovery note for non-firing PR workflows (pull_request event types + target branch validation).
  • Extended checkpoint criteria to include skill-driven iteration.

• Step 5 intro: add second predict-and-observe exercise

  • Added a short task-brief drafting exercise (predict → reveal → compare → revise) to improve mental-model formation before the 05b/05c sequence.
  • Updated checkpoint to include brief authoring/revision practice.

/agentic-workflows improve my PR reviewer brief so it highlights risky file types and keeps comments under 120 words

• Fixes [workshop-sim] Workshop Simulation Report — 2026-07-25 (Run #10, 1000×Monte Carlo) #1989

[pelikhan]

@copilot revise step 7 activity and checkpoint around the fact that the user needs to learn to always leverage the agent+skill to edit/improve/design agentic workflows

[Copilot]

@copilot revise step 7 activity and checkpoint around the fact that the user needs to learn to always leverage the agent+skill to edit/improve/design agentic workflows

Updated in 59219ff. Step 7d now uses a /agentic-workflows prompt as the model-access activity, and the Step 7d checkpoint now explicitly requires using the agent + skill to improve workflow design decisions.