[WIP] Update example READMEs for AI output files and diagram explanation

[Copilot]

• Update ai/examples/01_peered_vnets/README.md:
  • Add "## What This Example Produces" section (after title, before The Diagram)
  • Expand "What I Saw" with richer visual walkthrough
  • Add "### Step 5: The Two Outputs" inside "How I Translated It"
  • Add 5th lesson in "What I Learned" about README as second artifact
• Update ai/examples/02_distributed_app/README.md:
  • Add "## What This Example Produces" section
  • Expand "What I Saw" with richer visual walkthrough
  • Add "### Step 5: The Two Outputs" inside "How I Translated It"
  • Add 5th lesson in "What I Learned" about README as second artifact

Original prompt

Goal

Update the two example READMEs under ai/examples/ so that each one:

1. Explicitly produces two output files as part of the worked example:

   • The .tfvars file (already exists as output.tfvars)
   • A complete README.md (the example README itself, which should be described as also being an AI-generated output)

2. Showcases the original diagram image prominently, with the ![diagram](diagram.png) embed near the top.

3. Explains in detail how the diagram was interpreted — what the AI saw, what decisions it made, what it guessed vs. asked about, and how the relational model shaped the output.

---

Files to Update

ai/examples/01_peered_vnets/README.md

Current BlobSha: 29a70d542149c34077847681f10f00499fc5fe03

This README is already well-structured but needs the following changes:

1. Add a preamble section at the top (after the title, before "The Diagram") that describes both output files generated from the diagram:

   • app01.tfvars — the TFVARS file
   • README.md — this document, which is itself a generated artifact explaining the AI's interpretation

2. Keep the diagram embed (![Hand-drawn architecture diagram](diagram.png)) prominent.

3. Expand "What I Saw" to include a more detailed visual walkthrough — describe what specific marks, labels, boxes, and arrows are visible in the image and what they mean.

4. Expand "How I Translated It" to include a subsection called ### Step 5: The Two Outputs that explains:

   • The AI generated app01.tfvars containing the infrastructure declaration with # REVIEW: and # EXPLAIN comments
   • The AI also generated this README as a second artifact — a human-readable explanation of every decision, serving as the audit trail

5. Update the output reference to point to app01.tfvars (which it already does) and make clear that the README is the second output.

6. In "What I Learned", add a 5th lesson about generating the README as a second output artifact.

---

ai/examples/02_distributed_app/README.md

Current BlobSha: 2243fb7b6539afc1304a9b9274bb051b5fcec75f

Same pattern of changes:

1. Add a preamble section describing both output files:

   • az_app.tfvars — the TFVARS file
   • README.md — this document, the AI's interpretation narrative

2. Keep the diagram embed prominent.

3. Expand "What I Saw" with more visual detail about what specifically appears in the image.

4. Expand "How I Translated It" with a ### Step 5: The Two Outputs section explaining the dual-output nature.

5. In "What I Learned", add a lesson about the README as a second output.

---

Tone and Style

• These READMEs are written in first person from the AI's perspective ("I saw", "I translated", "I inferred").
• They explain reasoning, not just results.
• They are honest about uncertainty — what was guessed, what was asked, what was left for review.
• They treat the README itself as a deliverable, not just documentation.

---

Context

The existing file contents are as follows:

ai/examples/01_peered_vnets/README.md (current):

# Example: Two Peered VNets with Role-Based VMs

## The Diagram

![Hand-drawn architecture diagram](diagram.png)

A hand-drawn sketch on paper. No tooling, no structured format — just boxes, labels, and arrows.

## What I Saw

Reading the diagram left to right, top to bottom:

1. **An outer box labeled "Resource Group - App01"** — everything lives in a single resource group.
2. **AppVNet (10.100.0.0/16)** — the left VNet, explicitly labeled with its CIDR range, containing two subnets:
   - **Web Subnet** with 3 VMs: Web 1, Web 2, Web 3
   - **API Subnet** with 3 VMs: API 1, API 2, API 3
3. **DB VNet (10.200.0.0/16)** — the right VNet, also labeled with its CIDR, containing:
   - **DB Subnet** with 3 VMs: DB 1, DB 2, DB 3
4. **A dashed line labeled "Peered"** between the two VNets.
5. **Two NSG annotations:**
   - "NSG Can Access" arrow from **Web Subnet → DB Subnet**
   - "NSG Can Access" arrow from **API Subnet → DB Subnet**

## How I Translated It
... (full existing content)

ai/examples/02_distributed_app/README.md (current):

# Example: Availability Zone Distribution with Port-Specific NSGs and DR Peering

## The Diagram

![Hand-drawn architecture diagram](diagram.png)

A hand-drawn sketch showing a single VNet with three subnets, VMs distributed across two explicit availability zones, port-specific NSG annotations, and a peered DR mirror deployment.

## What I Saw
... (full existing content)

The full current content of both READMEs and both .tfvars files is available in the repository at the paths above.

---

Summary of Changes Required

For EACH of the two example READMEs:

1. Add a "## What This Example Produces" section near the top (after the title, before "The Diagram") listing both output files with descriptions.
2. Expand the diagram section with richer visual narration.
3. Add a **"##...

This pull request was created from Copilot chat.