home / skills / sandraschi / advanced-memory-mcp / code-review-assistant

code-review-assistant skill

This skill guides high-quality, maintainable code reviews by applying heuristics, automation, and best practices across modules, teams, and project goals.

npx playbooks add skill sandraschi/advanced-memory-mcp --skill code-review-assistant

Review the files below or copy the command above to add this skill to your agents.

Files (9)

SKILL.md

1.8 KB

---
name: code-review-assistant
description: Expert code reviewer focusing on quality, maintainability, performance, and best practices
license: Proprietary
---

# Code Review Assistant
> **Status**: ✅ Research complete
> **Last validated**: 2025-11-08
> **Confidence**: 🟡 Medium — Research-backed guidance – review quarterly

## How to use this skill
1. Triage change context via [modules/core-guidance.md](modules/core-guidance.md).
2. Apply review heuristics from [modules/review-techniques.md](modules/review-techniques.md).
3. Leverage automation described in [modules/automation-and-tooling.md](modules/automation-and-tooling.md).
4. Align team rituals and metrics using [modules/team-practices.md](modules/team-practices.md).
5. Track open research follow-ups in [modules/known-gaps.md](modules/known-gaps.md) and refresh quarterly with [modules/research-checklist.md](modules/research-checklist.md).

## Module overview
- [Core guidance](modules/core-guidance.md) — sets review goals, scope, and triage questions.
- [Review techniques](modules/review-techniques.md) — checklist, heuristics, language-specific tips.
- [Automation & tooling](modules/automation-and-tooling.md) — bots, static analysis, AI assistance, metrics.
- [Team practices](modules/team-practices.md) — guidelines, onboarding, continuous improvement.
- [Known gaps](modules/known-gaps.md) — outstanding experiments and deeper dives.
- [Research checklist](modules/research-checklist.md) — quarterly validation workflow.

## Research status
- Verified against Google/Microsoft guidance and 2024 GitHub review metrics.
- Next update due 2026-02-01 or sooner if new GenAI-assisted review policies emerge.
- Known gaps capture future experiments (e.g., AI suggestions, pair review metrics).*** End Patch

Overview

This skill is an expert code review assistant that focuses on quality, maintainability, performance, and best practices for Python projects. It guides reviewers through triage, targeted heuristics, and automation to produce consistent, actionable feedback. The skill also supports team alignment and continuous improvement for review rituals and metrics.

How this skill works

The assistant inspects change context, tests, and dependency impact to prioritize what needs human attention versus automated checks. It applies a compact checklist and language-specific heuristics to detect design smells, security issues, performance regressions, and maintainability risks. It recommends automation hooks (linting, static analysis, CI gates) and suggests measurable team practices for tracking review effectiveness.

When to use it

During pull request triage to prioritize risky changes
When assessing design and architecture impacts of code changes
Before merging performance-sensitive or security-related updates
To standardize review quality across a team
When onboarding reviewers or improving review metrics

Best practices

Triage by risk: evaluate scope, data flow, and external interfaces first
Use small, focused reviews to reduce cognitive load and accelerate feedback
Combine automated checks with focused human review for logic and design
Document recurring issues and convert them into linters or CI tests
Keep review comments actionable and link to examples or tests

Example use cases

Reviewing a Python service change that alters database schema and access patterns
Evaluating a refactor for code readability and test coverage impact
Catching potential performance regressions introduced by new dependencies
Designing CI checks that block common mistakes discovered during reviews
Setting up a team playbook and metrics for consistent reviewer behavior

FAQ

How quickly can this skill be applied in an existing repo?

You can start using the heuristics and triage flow immediately; automated tooling and CI integration typically take a few hours to days depending on existing pipelines.

Does this replace automated tools or human reviewers?

No. It augments both: it prescribes which checks to automate and which decisions require human judgment, improving overall review efficiency and quality.