One-click install
npx skills add https://github.com/nodatall/primedirective --skill merge-review-nodatall
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: merge-review
Source: https://github.com/nodatall/primedirective/tree/main/skills/merge-review
Command: npx skills add https://github.com/nodatall/primedirective --skill merge-review-nodatall

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Manual branch merge reviews are often inconsistent, miss critical issues, and lack structured tracking, leading to buggy code being merged or lengthy rework cycles after merge attempts.

Core Features & Use Cases

  • Structured Merge-Readiness Loop: Runs a repeatable review, fix, validate, rereview workflow against the origin/main...HEAD diff to catch correctness, structural, test, and production readiness issues.
  • Durable State Tracking: Records all findings, fixes, validation results, and blockers in a persistent state document so work can be resumed seamlessly across sessions.
  • Clear Merge Verdicts: Classifies issues as fixable, needing human decision, residual risk, or no action, and only marks the branch merge-ready when all fixable issues are resolved and validated.
  • Use Case: A software engineer preparing a feature branch for merge can run this skill to automatically identify and fix verified local issues, get a clear list of items requiring human input, and confirm the branch meets merge standards without ad-hoc manual review.

Quick Start

Use the merge-review skill to run a full merge-readiness check on your current feature branch, fix all verified local issues, and confirm it is ready to merge.

Frequently Asked Questions about merge-review

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

FAQPage Schema
How do I automate branch merge-readiness checks before merging a feature branch?

Automating branch merge-readiness checks involves running a structured review loop against the origin/main...HEAD diff to validate correctness, structural quality, test coverage, and production readiness before merging. This eliminates inconsistent manual pre-merge checks and ensures feature branches meet merge standards.

What is a merge-readiness review loop in a git workflow?

A merge-readiness review loop in a git workflow is a repeatable process of reviewing, fixing, validating, and re-reviewing feature branches against the base main branch. It tracks findings, fixes, and validation results in a persistent state document to ensure production readiness before merging.

Can I resume an incomplete code review across sessions if the branch validation is interrupted?

Yes, you can resume an incomplete code review across sessions using durable state tracking. This records all findings, fixes, validation results, and blockers in a persistent state document, allowing work to be resumed seamlessly without losing prior review progress on the feature branch.

How are code review findings classified when preparing a branch for merge?

Code review findings are classified as fixable, needing human decision, residual risk, or no action when preparing a branch for merge. A branch is only marked merge-ready when all fixable issues are resolved and validated, with any remaining items clearly reported as residual risk.

Does manual pre-merge code review miss critical issues compared to an automated approach?

Manual pre-merge code review often misses critical issues and lacks structured tracking, leading to buggy code being merged or lengthy rework cycles. An automated approach applies consistent validation for correctness, test coverage, and production readiness, providing clear merge-readiness verdicts with residual risk reporting.

What are the limitations of automating merge review for feature branches?

Automating merge review for feature branches cannot resolve issues requiring human decision, which are classified separately as blockers. The automation only marks a branch merge-ready when all fixable issues are resolved and validated, leaving residual risk and items needing human input explicitly reported for manual intervention.