What documents does the review agent analyze?

It searches the .claude/pipeline directory for design and specification markdown files, as well as the service profile for platform requirements.

Does this skill work with any Git branch?

Yes, it can review changes on the current branch, a specific feature branch, or any diff relative to a specified base branch.

Where are the review results stored?

Findings are saved to .claude/review-findings.jsonl, providing a persistent and machine-readable record of the review output.

How does the skill identify which feature to review?

The skill automatically detects the feature ID by parsing the current Git branch name, or you can manually provide a feature ID as an argument.

Can I use this skill for general code quality checks?

Yes, by using the --diff flag, you can perform a general code quality review against a base branch even without specific pipeline artifacts.

Datum Cloud Code Review

Name: Datum Cloud Code Review
Author: datum-cloud

bydatum-cloud

•

Collaboration Tools

Automates high-context code reviews by correlating Git changes with design artifacts and architectural specifications.

The Datum Cloud Review skill empowers Claude to perform intelligent, context-aware code reviews within your development pipeline. By automatically detecting feature IDs from Git branches, it pulls relevant design documents, technical specs, and service profiles to ensure that code changes align with the intended architecture. It leverages a specialized code-reviewer agent to identify blocking issues, warnings, and improvements, logging findings in a structured format for seamless integration into your CI/CD or approval workflows.

Key Features

01Architecture validation against design artifacts and specifications

02Platform-specific checks for telemetry, quotas, and activity requirements

03Structured output of findings to JSONL for pattern analysis

04Automatic branch-to-feature mapping for context loading

05Flexible review scopes including specific feature IDs or branch diffs

062 GitHub stars

Use Cases

01Pre-submission code audits to catch blocking issues early

02Comparing local branch changes against a base branch for quality assurance

03Validating feature implementations against original architectural designs

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add datum-cloud/claude-code-plugins review

For use in Claude.ai and ChatGPT

Download Skill