Add --dry-run Mode to Simulate Playbook Execution

[shalev007]

Problem:
Currently, there's no way to preview what a playbook would do without actually making HTTP requests. This makes debugging risky and error-prone—especially in destructive workflows (e.g., POSTing data or deleting resources).

A dry-run mode would let developers inspect:

• Resolved variables
• Final request URLs
• Headers and body content (after Jinja rendering)

---

✅ What to Implement

• Add a --dry-run flag to the CLI (restbook run --dry-run)
• Simulate each step without sending requests
• Print:
  • Resolved endpoint
  • Final headers
  • Final body / payload (from data or fromFile)
  • Stored variable names
• Highlight any missing or unresolved variables
• Exit cleanly with exit code 0 (unless validation fails)

---

💡 Why It Matters

• Safer debugging during development
• Great for CI/CD dry runs or change previews
• Helps catch templating bugs early (missing vars, incorrect formatting)
• Reduces accidental API spam / destructive calls

---

📎 Related Enhancements

• Could later support a --dry-run=phase-2 or --until=phase-N for partial simulation
• Option to write dry-run output to a file (--dry-run-output out.json)
• Can integrate with restbook lint in the future