check

Review diffs, apply safe fixes, and run verification before sign-off.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/treejh/ai-agent-skills --skill check-treejh
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: check
Source: https://github.com/treejh/ai-agent-skills/tree/main/skills/check
Command: npx skills add https://github.com/treejh/ai-agent-skills --skill check-treejh

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Merge-ready code changes and pull requests often contain overlooked typos, missing imports, risky dependency changes, or security and architectural regressions that slow releases and introduce risk. This Skill reads diffs, applies unambiguous safe fixes, surfaces gated or manual items for human approval, and runs specialist reviews and verification so reviewers can trust the sign-off.

Core Features & Use Cases

  • Diff classification: Automatically measure the diff and choose Quick, Standard, or Deep review depth.
  • Autofix routing: Apply safe_auto fixes immediately, batch gated_auto fixes for confirmation, and surface manual items for sign-off.
  • Specialist orchestration: Launch security and architecture reviewers in parallel for Standard/Deep diffs and merge findings with severity rules.
  • Verification: Run the project's verification command and require passing results before declaring done.
  • Use Case: Run before merging a feature branch to auto-correct trivial issues, detect scope drift, run experts on risky changes, and produce a verification-backed sign-off.

Quick Start

Run the check skill on the current branch diff to auto-apply safe fixes, run specialists for deeper review as needed, and produce a verification-backed sign-off.

Frequently Asked Questions about check

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

FAQPage Schema
How do I automate code review for pull requests?▼

Automating code review for pull requests involves reading the diff, applying safe autofixes to trivial issues, batching gated fixes for confirmation, and running specialist security and architecture reviews in parallel.

What is safe autofix in diff analysis?▼

Safe autofix in diff analysis automatically applies unambiguous corrections to code changes immediately, while routing risky or gated modifications into batches that require explicit human approval before sign-off.

How does diff classification determine review depth?▼

Diff classification determines review depth by measuring the size and scope of code changes, automatically routing the diff into Quick, Standard, or Deep review modes to balance speed and thoroughness.

Can I run security and architecture reviews in parallel on local branch diffs?▼

Yes, you can run security and architecture reviews in parallel on local branch diffs. The system launches specialist reviewers concurrently for Standard and Deep modes, merging findings with severity rules.

When do I need a verification-backed sign-off for code changes?▼

You need a verification-backed sign-off when preparing merge-ready code changes. The process runs the project's verification command and requires passing results before declaring the review complete and safe to merge.

What are the limitations of automated diff review?▼

Automated diff review cannot fix complex or ambiguous issues automatically; it surfaces these as manual items for human sign-off. It also requires a configured project verification command to execute before completion.