Can it post comments directly to GitLab or Jira?

Yes, after presenting the review report, the skill offers an optional write-back step to post findings as comments to the relevant system via CLI or MCP.

Does it support parallel reviews?

Yes, it uses a sub-agent architecture to spawn multiple analysis tasks simultaneously, allowing for thorough reviews of large documentation sets or complex code changes.

How does the confidence score work?

The skill calculates a weighted average (0-100) across five dimensions: completeness (25%), quality (25%), accuracy (25%), ambiguity risk (15%), and open questions (10%).

Can it check for project-specific coding standards?

Yes, it automatically scans for CLAUDE.md, linter configurations, and project documentation to contextualize its feedback according to your local conventions.

What external tools are required for this skill?

It prefers the glab (GitLab) and acli (Atlassian) CLI tools for efficiency, but can fall back to standard MCP servers if the CLIs are unavailable.

AI-DLC Reviewer

Name: AI-DLC Reviewer
Author: onesmartguy

byonesmartguy

0•

协作工具

Performs senior-level peer reviews of GitLab merge requests and AI-DLC documentation artifacts with automated confidence scoring.

AI-DLC Reviewer is a specialized Claude Code skill designed to act as a senior engineer during the software development lifecycle. It automates the review process for both implementation work (GitLab Merge Requests) and documentation (Confluence and Jira artifacts), ensuring alignment between intent and execution. By spawning specialized sub-agents, it provides deep analysis, identifies inconsistencies, flags ambiguity risks, and generates comprehensive reports with weighted confidence scores to help teams maintain high-quality standards across their technical ecosystem.

主要功能

01Sub-agent driven analysis for parallel, multi-dimensional review processing

02Documentation validation for Confluence pages and Jira stories/tasks

03Quantitative confidence scoring based on completeness, quality, and accuracy

04Automated GitLab Merge Request (MR) peer review with deep code diff analysis

050 GitHub stars

06Direct integration with glab and acli for seamless command-line execution

使用场景

01Reviewing complex Confluence documentation sets for internal consistency and technical gaps

02Validating that a Jira Bolt (story) implementation matches the original Acceptance Criteria

03Conducting rigorous senior-level code reviews on GitLab MRs to identify edge cases and risks

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add onesmartguy/raptor-claude-plugins aidlc-review

For use in Claude.ai and ChatGPT

Download Skill