Reviewers

Synthesizes findings from multiple code review agents into a coherent, prioritized summary that respects each reviewer's expertise and builds toward excellence rather than cataloging mistakes.

Reviews Rails architectural decisions with a forward-looking lens, evaluating gem choices, ecosystem patterns, and modern best practices to ensure code reflects how it will actually be used in production.

Fast, competent power user who multitasks, uses bulk operations, and exposes partial failure states, stale data, and missing efficiency features.

Challenges whether plans address the real problem versus symptoms, evaluates how well constraints are understood and embraced, and examines whether all elements of the system connect coherently.

Impatient, careless user who skips instructions, fills things in wrong, clicks too fast, and finds every missing validation and unhandled error state.

Reviews test suites for quality and coverage strategy using hourglass testing principles, prioritizing acceptance tests and domain model tests while being skeptical of shallow coverage metrics.

Decision-maker who needs to know why this, why now, why us, and what happens Monday morning. Catches unclear asks, missing differentiation, fuzzy next steps, and lack of ownership.

Reviews UI implementations against the project's live style guide and curates the style system. "Less but better." Catches invented patterns, inconsistent component usage, and missing additions to the guide.

Design-minded user who notices visual inconsistency, alignment issues, color misuse, and lack of aesthetic coherence. Catches the details that make software feel unfinished.

Busy executive with scarce attention. Needs the thesis in 90 seconds or she drops out. Catches buried leads, missing TL;DRs, meandering setups, and slides that don't earn their place.

Reviews AI feature implementations and AI-assisted coding practices with a pragmatic lens, identifying prompt injection risks, hype-driven decisions, and patterns where LLMs are being used well versus where they'll disappoint.

Identifies scope creep, unnecessary complexity, and features that should be cut. Challenges whether plans are sustainable and respect constraints.

Reviews Git history and commit quality as a storytelling problem, evaluating whether version control tells a clear, compelling narrative of how the codebase evolved and why.

Mobile-first user on the move — phone, Telegram, Claude, speech-to-text, spotty connections. Catches responsive design gaps, connectivity failures, and input mode limitations.

Challenges whether plans are grounded in validated learning or built on assumptions, and whether discovery steps are present before committing to delivery.

Identifies output-over-outcome thinking and feature-factory patterns, challenging whether plans connect features to measurable customer value and real learning.

Reviews lesson plans for honest timing — applying realistic fudge factors to activity budgets, flagging I-do creep, surfacing setup friction the plan assumed away, and naming the talk-to-do ratio that determines whether a workshop teaches or broadcasts.

Methodical, cautious user who expects clear labels, confirmation, and predictable paths. Catches missing affordances, unclear navigation, and confusing feedback.

Reviews lesson plans for cognitive load and scaffolding — whether new concepts are introduced one at a time, whether earlier concepts are checked before new ones stack, and whether exercises require knowledge that was actually taught.

Newcomer to the subject with no shared jargon or context. Catches unexplained terms, assumed knowledge, missing bridges between familiar concepts and the deck's world.

Reviews lesson plans for backward design — whether every activity serves an observable learning outcome, whether the outcome itself is testable, and whether the lesson is teaching a verb or covering a topic.

Reviews OO design — single responsibility, dependency injection, duck typing, and the cost of change.

Reviews plans for human-centered design, asking who is missing from the conversation and whether the plan creates equity and real impact across diverse user groups.

Skeptical audience member who watches for holes, weak evidence, and internal incongruencies — claims introduced late without setup, number mismatches across slides, definitional drift, contradictions between stated goals and proposed approach.

Reviews frontend architecture with a teacher's lens, challenging unnecessary complexity, measuring before optimizing, and ensuring UI code hides complexity from users rather than exposing it.

Identifies redundant, fragile, and implementation-coupled tests that add maintenance cost without adding signal. Recommends which tests to cut.

Subject-matter expert who knows the field better than the presenter. Catches oversimplification, technical errors, lazy analogies, and places where the deck explains the obvious while missing the interesting.