review-tier

Classify pull-request change surfaces into T1, T2, or T3 review tiers from git evidence.

6|1|Updated Jul 11, 2026
One-click install
npx skills add https://github.com/jl-cmd/claude-dev-env --skill review-tier-jl-cmd
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: review-tier
Source: https://github.com/jl-cmd/claude-dev-env/tree/main/skill-archive/review-tier
Command: npx skills add https://github.com/jl-cmd/claude-dev-env --skill review-tier-jl-cmd

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Teams need a consistent, deterministic way to decide how much review rigor a change requires. This Skill inspects the current git change surface and classifies it into a T1, T2, or T3 review tier using a versioned policy, removing subjective guesswork from review routing. ## Core Features & Use Cases - Deterministic Tier Classification: Builds a hashed inventory of committed, staged, unstaged, and untracked changes, then scores bounded axes (packages, risk, public_api, dependencies) against policy thresholds. - Hard Trigger Escalation: Automatically forces T3 when changes touch security, migration, public-api, or release markers. - Override Governance: Applies explicit tier overrides, permitting upgrades and only policy-approved downgrades, and produces an immutable decision card with policy and diff hashes. - Use Case: Before opening a pull request, run the classifier to learn that a change touching migrations/ is T3 and must route through the strictest review path via /review-router. ## Quick Start Ask the AI to classify the current repository changes with /review-tier and report the calculated and effective review tier.

Frequently Asked Questions about review-tier

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

FAQPage Schema
How do I classify a pull request into review tiers automatically?▼

Run the review-tier classifier against the repository; it builds a canonical inventory of all changed files from git, scores bounded axes like packages and risk, and returns a calculated T1, T2, or T3 tier plus an effective tier after any override.

What triggers a T3 review tier classification?▼

A T3 tier results from hard triggers such as security, migration, public-api, or release markers in changed paths, or when the weighted axis score meets the T3 threshold of 5 in the tier policy.

Can I override the calculated review tier?▼

Yes, explicit overrides are supported. Upgrades to a higher tier are always allowed, while downgrades succeed only if the requested tier appears in the policy's approved_downgrades list; otherwise an UNAPPROVED_TIER_DOWNGRADE error is raised.

Does the review tier classifier require external Python packages?▼

No, the classifier uses only the Python 3.11+ standard library plus the git command-line tool. A non-zero script exit is treated as a failed classification.

Why does review-tier classification fail with an invalid base ref error?▼

The classifier raises INVALID_BASE_REF when the requested base reference cannot be verified by git rev-parse. Without an explicit base, it falls back to the remote HEAD or the main branch, and raises AMBIGUOUS_BASE_REF if neither resolves.