alignment-review

Assess code changes for alignment with OpenEnv principles and RFCs.

2.5k|421|Updated Oct 1, 2025
One-click install
npx skills add https://github.com/meta-pytorch/OpenEnv --skill alignment-review
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: alignment-review
Source: https://github.com/meta-pytorch/OpenEnv/tree/main/.claude/skills/alignment-review
Command: npx skills add https://github.com/meta-pytorch/OpenEnv --skill alignment-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a structured, automated approach to reviewing code changes for alignment with OpenEnv principles and RFCs. It helps teams catch deviations before merging by standardizing how reviews are performed.

Core Features & Use Cases

  • Automated checks: Run lint and debug hooks to surface issues early (Tier 1).
  • RFC-driven review: Scan RFCs to identify conflicts or open design discussions affecting the changes.
  • Two-tier review workflow: Distinguish immediate fixes (Tier 1) from alignment discussions (Tier 2), guiding reviewers toward actionable outcomes.
  • Contextual guidance: Use documented principles and invariants to validate changes during PR reviews, pre-commit checks, or change audits.

Quick Start

Run the automated checks and read the policy docs to begin a review:

  • Run: bash .claude/hooks/lint.sh and bash .claude/hooks/check-debug.sh
  • Read: .claude/docs/PRINCIPLES.md and .claude/docs/INVARIANTS.md
  • Inspect RFCs: scan the rfcs/ directory for status and notes
  • If issues arise, categorize them as Tier 1 or Tier 2 during your review and prepare a concise report.

Frequently Asked Questions about alignment-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review code changes for alignment with project principles?

Code alignment review assesses whether changes conform to documented principles and RFCs. Run automated lint and debug checks first, then scan RFCs for conflicts, and categorize issues into Tier 1 (immediate fixes) or Tier 2 (design discussions) to guide structured review outcomes.

What's the difference between Tier 1 and Tier 2 issues in code review?

Tier 1 issues are immediate, actionable fixes caught by automated lint and debug checks. Tier 2 issues require alignment discussions tied to RFCs or principles. This two-tier model helps reviewers prioritize and separate quick corrections from deeper design conversations.

Can I automate code reviews against OpenEnv principles?

Yes. Automated checks via lint and debug hooks surface Tier 1 issues during PR reviews or pre-commit checks. Combine these with manual RFC scanning and principle validation to enforce alignment before merging, standardizing your review process across the repository.

When should I run alignment checks—pre-commit or during PR review?

Alignment checks work at both stages. Run lint and debug hooks as pre-commit checks to catch issues early, or apply them during PR review for broader audits. The two-tier workflow supports either timing, letting teams choose when to surface Tier 1 and Tier 2 issues.

What do I need to set up before running alignment reviews?

Review the documented principles and invariants in your repository, ensure RFCs are organized in the rfcs/ directory with status notes, and configure lint and debug hook scripts. These prerequisites prepare your codebase for structured, principle-driven reviews.

How do RFCs factor into code alignment decisions?

RFCs represent design decisions and open discussions. During review, scan them to identify conflicts or design implications affecting the changes. This ensures code modifications respect approved designs and flag areas where RFC-level decisions are needed before merging.