Add THE CANONICAL 100: Complete Lucidia language definition through examples

This commit introduces the foundational specification for Lucidia v1.0 - a set
of 100 working example programs that DEFINE the language through demonstration
rather than formal grammar.

Key Philosophy:
- Examples ARE the spec (not documentation OF the spec)
- AI systems learn by reading all 100 examples and extracting patterns
- Humans learn by working through examples sequentially
- No feature exists unless demonstrated in these examples

Structure:
- 001-010: Fundamentals (hello world → functions)
- 011-020: Data & Collections (lists, maps, sets)
- 021-030: Control Flow (if, loops, pattern matching)
- 031-040: Functions & Composition (map, filter, reduce, closures)
- 041-050: UI Basics (forms, inputs, validation)
- 051-060: Reactive Programming (state, watchers, events)
- 061-070: Consent & Privacy (permission system - CORE DIFFERENTIATOR)
- 071-080: Storage & Sync (local-first, cloud-optional)
- 081-090: AI Integration (intent → code, learning user style)
- 091-100: Complete Applications (todo, notes, chat, e-commerce)

Core Language Features Demonstrated:
✓ Intent over ceremony (write WHAT, not HOW)
✓ Consent as syntax (ask permission for: resource)
✓ Local-first storage (store locally, sync to cloud optional)
✓ AI-collaborative (### Intent comments become code)
✓ Reactive by default (state, watch, computed)
✓ Zero setup (runs in browser via WASM)
✓ Multi-paradigm (functional, OOP, reactive, agent-based)
✓ Gradual complexity (hello world → production apps)

Files Created:
- README.md - Learning philosophy and path
- INDEX.md - Complete reference table
- 001-100.lucidia - All example programs

Total: 102 files, ~3,500+ lines of example code

Why This Matters:
This is not just documentation. This IS Lucidia. Every parser, compiler,
AI assistant, and developer tool will be trained on these examples. They
are the permanent, immutable foundation of the language.

Next Steps:
1. Build parser that learns from these examples
2. Train AI to recognize and generate Lucidia patterns
3. Create browser playground with these as gallery
4. Use for academic paper and conference presentations

Designed by: Cece (Principal Language & Runtime Architect)
For: BlackRoad Operating System / Lucidia Programming Language
Status: Complete foundation for implementation

docs/examples/canonical/083-ai-text-summary.lucidia

@@ -0,0 +1,41 @@
+# 083: AI Text Summary
+# Automatic summarization
+
+# Long text
+article = """
+Lucidia is a revolutionary programming language designed for the modern era.
+Unlike traditional languages that were built for compilers, Lucidia is built
+for intelligences - both human and artificial. It features consent as syntax,
+local-first storage, and seamless AI collaboration. The language runs natively
+in browsers via WebAssembly, requiring zero setup. Users can write intent
+instead of implementation, letting AI handle the ceremony while maintaining
+human creative control.
+"""
+
+# Summarize
+summary = ai.summarize(article, max_words: 20)
+show summary
+# Output: "Lucidia is a modern programming language with AI collaboration,
+#          consent-as-syntax, and browser-native execution."
+
+# Bullet point summary
+logs = load "server_logs" locally
+
+### Intent: Summarize last 24 hours of logs as bullet points
+### Group by severity, show counts
+
+bullet_summary = ai.summarize(logs, format: "bullets")
+show bullet_summary
+
+# Expected:
+# • Errors: 3 (database timeout, API failure, memory leak)
+# • Warnings: 47 (slow queries, deprecated API usage)
+# • Info: 1,284 (normal operations)
+
+# Different summary styles
+ai.summarize(data, style: "executive")   # High-level overview
+ai.summarize(data, style: "technical")   # Detailed technical
+ai.summarize(data, style: "eli5")        # Explain like I'm 5
+
+# AI picks the important parts
+# You don't have to