background-reviewerlisted

# Background Reviewer Multi-LLM parallel code review with specialized agent personas. ## Quick Start ```bash # 1. Determine review scope and round number # 2. Create: .context/reviews/ # 3. Run review agents with personas in background # 4. Each persona saves findings to {round}-{persona}.md # 5. Merge into prioritized findings when ready ``` ## Output Convention ``` .context/reviews/ ├── R01-security-reviewer.md # Round 1: security persona ├── R01-architecture-reviewer.md # Round 1: architecture persona ├── R01-code-quality-reviewer.md # Round 1: code quality persona ├── R01-performance-reviewer.md # Round 1: performance persona ├── R01-merged.md # Round 1: merged findings ├── R02-security-reviewer.md # Round 2: after fixes, re-review └── R02-merged.md ``` **Round number** increments each time reviews are run (R01 = first pass, R02 = after fixes, R03 = final check). Detect next round: ```bash ROUND=$(printf "R%02d" $(( $(ls .context/reviews/R*-*.md 2>/dev/null | sed 's/.*\/R\([0-9]*\)-.*/\1/' | sort -rn | head -1 | sed 's/^0*//') + 1 ))) # → R01, R02, R03, ... ``` ## Persona-Based Review (Recommended) Each reviewer adopts a detailed agent persona from `.agents/personas/` or `~/.agents/personas/`. Personas define identity, review lens, evaluation framework, and output format — producing consistent, specialized reviews. ### Available Personas ```bash agt persona list # See all available personas agt person