docs(agent-development): add minimal working example after Overview

[sjnims]

Summary

Adds a Quick Start section with a minimal, copy-paste ready agent example to improve onboarding for new users. Also enhances the model field documentation to more clearly emphasize inherit as the recommended default.

Problem

Fixes #25

The first example users see is the "Complete Format" which includes all frontmatter fields with explanations. New users often want a quick-start template that "just works" before diving into all the options.

Solution

Added a Quick Start section immediately after Overview with:

• ~25-line minimal working agent example
• Complete and copy-paste ready
• Link to detailed Agent File Structure section for more options

Also improved the model field recommendation:

**Recommendation:** Use `inherit` (recommended default) unless the agent specifically needs:
- `haiku` for fast, cost-sensitive operations
- `opus` for complex reasoning requiring maximum capability

Alternatives Considered

• Putting Quick Start at the end: Rejected because new users read top-to-bottom
• Making the Complete Format example smaller: Would lose important context for advanced users

Changes

• plugins/plugin-dev/skills/agent-development/SKILL.md: Added Quick Start section, improved model recommendation

Testing

• Markdownlint passes
• Word count remains under 2,000 (1,891 words)
• Example is complete and valid agent format

---

🤖 Generated with Claude Code

[claude]

Claude finished @sjnims's task —— View job

---

Review Summary ✅

Status: Approved - this is a high-quality improvement to the documentation.

---

What's Good

1. Excellent Placement
The Quick Start section is positioned perfectly - right after the Overview and before diving into component comparisons. This gives new users an immediate "hello world" they can copy-paste and experiment with.

2. Well-Crafted Example
The minimal agent example (lines 23-51) is genuinely minimal yet complete:

• ~29 lines of actual content
• Includes all required frontmatter fields (name, description, model, color)
• Contains a proper <example> block with context, user request, assistant response, and commentary
• System prompt is concise but structured (role, process, output)
• Uses recommended defaults (inherit model, blue color)

3. Clear Progressive Disclosure
The link to "Agent File Structure" on line 53 provides a clear path for users who need more options, following the progressive disclosure pattern mentioned in CLAUDE.md.

4. Improved Model Guidance
The enhanced model field recommendation (lines 201-203) is much clearer:

• Explicitly states inherit as "recommended default"
• Provides specific criteria for when to deviate (haiku/opus)
• Removes ambiguity from the previous version

5. Quality Metrics Met

• ✅ Markdownlint passes (no issues)
• ✅ Word count remains under 2,000 limit (1,891 words as stated)
• ✅ Valid agent format
• ✅ Follows skill structure conventions

---

Suggestions

None - this PR is ready to merge as-is.

Minor observations (not blockers):

• The example uses my-reviewer as the agent name, which is appropriately generic for a tutorial
• The code review example aligns well with one of the plugin's actual agents (plugin-validator), providing good context continuity

---

Checklist Verification

From the PR description, all items confirmed:

• ✅ Markdownlint passes (verified)
• ✅ Word count under 2,000 (1,891 words)
• ✅ Example is complete and valid agent format (verified)
• ✅ Addresses issue docs(agent-development): add minimal working example after Overview #25 (adds minimal working example)
• ✅ Improves model field documentation (clearer recommendations)

---

Architecture Compliance

This change aligns with repository conventions from CLAUDE.md:

• ✅ Skill body uses imperative/infinitive form
• ✅ Progressive disclosure pattern (basic → detailed)
• ✅ Word count within skill guidelines (1,500-2,000 words)
• ✅ Markdown linting standards met

---

Recommendation: Merge this PR. It significantly improves the onboarding experience for new users creating their first agent while maintaining all quality standards.