What problem does it solve? Getting useful feedback on code, architecture, or experiments often requires a domain expert's perspective, which is not always available. This Skill runs a structured review of your current work through the lens of a specialized persona defined in an agents/personas directory, producing severity-ordered findings. ## Core Features & Use Cases - Persona-Based Review: Reads a selected persona file and evaluates your changed files, tests, docs, and config against that persona's standards for correctness, architecture, performance, and observability. - Automatic Persona Routing: Resolves the ai_tools root, discovers personas under agents/personas/, and routes by task intent using router.md files when no persona is explicitly named. - Structured Findings Output: Returns the persona used, findings ordered by severity with impact and recommended fixes, open questions, and suggested next actions. - Use Case: After refactoring a data pipeline, ask for a review as a data engineering expert and receive concrete, evidence-based critique scoped to your actual changes. ## Quick Start Review my current changes as a backend architect persona and list the findings by severity.
