Expand agent config coverage and refresh docs

Remove slop words from notes and examples

Add note 21 (multi-agent security), examples 06-07, restructure note 20

Revert "Add note 19: LLM Tokenization Deep Dive"

Add note 19: LLM Tokenization Deep Dive

Add notes 18-19 (token optimization, resource attacks) and update note 14

Renumber note 18 to 17 (Unicode Variation Selector Attacks)

Expand pattern database to 151 patterns across 15 categories

Add test suite, severity calibration guide, advisory hook docs

Add note 18: Unicode Variation Selector attacks on LLMs

Remove note 17 cross-references (not yet published)

Style: tighten narration and slop in notes 13, 16

Style: trim prose and slop in notes 01, 02, 05, 11

Style: minor slop/hedging fixes in notes 06, 07, 08, 12, 15

Consolidate notes: merge 01+04, renumber 00-16, un-defang source URLs