Add AGENTS.md for AI coding agent guidance

[Copilot]

Creates operational documentation for AI coding agents working in this GitHub Copilot workshop repository.

Changes

• AGENTS.md (181 lines) — Comprehensive agent instructions including:

  • Executable commands for development (uvicorn src.app:app --reload), testing (pytest), and validation
  • Code style patterns with FastAPI-specific examples (type hints, HTTPException usage, in-memory storage constraint)
  • Three-tier boundary system (✅ safe / ⚠️ ask / 🚫 never) protecting workshop infrastructure
  • Workshop context explaining 5-step learning progression and automated validation
  • NOS Culture alignment mapping workshop goals to organizational principles

• CLAUDE.md — Symlink to AGENTS.md for Claude-based agent compatibility

Key Constraints Documented

# ✅ Always safe - FastAPI validation pattern
def signup_for_activity(activity_name: str, email: str):
    if activity_name not in activities:
        raise HTTPException(status_code=404, detail="Activity not found")
    if email in activity["participants"]:
        raise HTTPException(status_code=400, detail="Student is already signed up")

# 🚫 Never - in-memory storage is intentional for workshop simplicity
# Don't add database layers or persistence

Workshop-specific boundaries prevent agents from modifying .github/workflows/ (breaks step validation), .github/steps/ (instructional content), or replacing in-memory storage (pedagogical choice).

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.