Skip to content

[workshop-sim] Repair: Name the exact AI-agent surface in Step 7's authoring prompt #3479

Description

@github-actions

Problem statement

Step 7 (07-your-first-workflow.md) tells learners to "run this prompt" in your AI agent without naming a specific surface (Copilot CLI vs. VS Code Copilot Chat vs. the Agents-tab prompt box). The Monte Carlo simulation shows this step is now the top dropout driver (23.5% conditional dropout among 22,269 at-risk runs), split between copilot-access-missing (3,762 failures) and workflow-authoring-friction (1,471 failures). tool: CCA learners are disproportionately affected because they must independently translate the generic phrase into the Agents-tab flow.

Proposed change

Add one short, explicit branch immediately before the authoring prompt block that names the exact surface per tool preference (Copilot CLI command, VS Code Copilot Chat panel, or GitHub Agents tab), mirroring the pattern already used correctly in the earlier "Verify Copilot access" section of the same file (which does say "In the terminal that is already open in your Codespace, run: gh copilot").

Failure mode classification

Access barrier — the ambiguity is about which tool surface to use, not about a missing concept.

Quantitative guardrail

  • Current overall_score for 07-your-first-workflow.md: not in the bottom-8 list (estimated mid-range, ~6.0/10 based on corpus mean); this repair should not lower it.
  • Weakest rubric dimension: checkpoint_quality (0.0, shared rubric-wide artifact) — unaffected by this change.
  • Current learning KPI dimensions for this step: active_learning and scaffolding should stay flat or improve, since the change adds a concrete decision point (a scaffolding aid) without removing any existing practice.
  • This is a clarifying addition, not a simplification of content — it should not reduce cognitive load below what the step already requires.

Acceptance criteria

  • The shared curriculum quantitative assessment overall_score for 07-your-first-workflow.md stays flat or improves after the change
  • The step's learning KPI index (2.0 × active_learning + 2.0 × checkpoint_quality + 1.5 × scaffolding) / 5.5 stays flat or improves
  • Hands-on practice, checkpoints, and scaffolding (the existing "Before You Start" checklist and final Checkpoint) are preserved or strengthened, not removed or weakened
  • The Monte Carlo simulator's copilot-access-missing and workflow-authoring-friction failure counts for 07-first-workflow decrease in a rerun

Suggested owner

workshop maintainer (content edit with copilot coding agent assist for wording)
Related to #3478

Generated by 🔬 Workshop Student Simulator · copilot · auto · 131.4 AIC · ⌖ 7.67 AIC · ⊞ 15.2K ·

  • expires on Sep 18, 2026, 8:54 AM UTC

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions