Optimize review-user-guide, review-architecture, analyze-db skills (~75% smaller)

[ydesgagn]

Summary

Applies the same optimization pattern proven on code-review-deep to the three largest review skills. Each was bloated with the same anti-patterns: a multi-page inline checkpoint log, per-framework duplicated detection commands, and 4-8 inline templates the model already knows how to render.

Key changes:

• review-user-guide (1460 → 328 lines, -77%) — replaced the 56-line inline checkpoint tracker with TaskCreate-based phase tracking; consolidated per-framework detection commands into compact bullet lists; replaced the four huge inline templates (Web/CLI/API/Library, ~620 lines) with structured per-product-type guidance that lists required sections and key formatting rules. Preserved all behavioral rules: walkthrough-style step-by-step guides, conditional visibility callouts, UI help-text reuse.
• review-architecture (1459 → 393 lines, -73%) — same checkpoint-log replacement; flattened the 95-line exempt-type detection into a single decision table; merged Standard and ML/DL templates (~350 lines combined) into compact per-section bullets. Preserved the SOUP validation logic (Requirements / Risk Level / Verification Reasoning), the IEC 62304 risk-level table, and the case-insensitive file checks.
• analyze-db (1392 → 330 lines, -76%) — replaced 184 lines of per-framework detection patterns with a single consolidated table; collapsed 257 lines of repetitive per-DB connectivity / schema queries into a single CLI Command Reference table; merged 5+ inline templates (SQL/MongoDB/Elasticsearch/Redis/BigQuery/Multi-DB) into per-DB section requirement bullets. Preserved every safety rule: large-table safeguards (estimated counts, sampling, no COUNT(*) on large tables), partition-filter requirements for BigQuery, anti-pattern lists, and the must-document-every-table mandate.

Total: 4311 → 1051 lines (-76%) across the three files, with every quality-preserving rule kept intact.

Types of changes

• Bugfix (fixes an issue)
• New feature (adds functionality)
• Refactoring (improves code without changing functionality)
• Breaking change (incompatible changes)
• Build or security update (updates dependencies, libraries, or security patches)
• Code style or documentation update (formatting, renaming, or documentation changes)
• Other (please describe):

Checklist

• Unit tests added to validate my fix/feature
• I have manually tested my change
• I did not add automation test. Why ?: Skill files are markdown prompts with no executable code path. Verified by running linters (markdown, GitHub Actions, YAML) — all 0 errors.
• Database changes requiring migration with downtime or reprocessing of existing data
• The SOUP file lists the risk Level, requirements and verification reasoning associated with each library
• readme.md includes sections on introduction, installation, usage, and contributing
• docs/architecture.md includes sections on the architecture diagram, software units, software of unknown provenance, critical algorithms and risk controls related to PII and security
• Impact on PII, privacy regulations (CCPA/GDPR/PIPEDA), CIS benchmarks or security (availability/confidentiality/integrity); management must be notified