forge-review

Orchestrate multi-agent code reviews and generate severity-tiered Markdown reports.

3|2|Updated Apr 3, 2026
One-click install
npx skills add https://github.com/fluxforgeai/ARTEMIS --skill forge-review
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: forge-review
Source: https://github.com/fluxforgeai/ARTEMIS/tree/main/.claude/skills/forge-review
Command: npx skills add https://github.com/fluxforgeai/ARTEMIS --skill forge-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The forge-review skill orchestrates a coordinated, read-only code review across a repository to detect algorithmic complexity issues, inappropriate data structures, paradigm inconsistencies, and performance anti-patterns, removing the manual overhead of deep, multi-dimensional reviews.

Core Features & Use Cases

  • Multi-agent orchestration: Launches four specialist subagents (Complexity Analyst, DS&A Reviewer, Paradigm Enforcer, Efficiency Sentinel) in parallel and aggregates structured findings.
  • Deduplication & severity tiers: Merges overlapping findings, assigns Critical/Warning/Suggestion tiers, and numbers items for clear triage.
  • Pipeline integration: Reads loop state for lifecycle handoff, writes structured Markdown reports to docs/reviews/, and enforces read-only guarantees on source files.
  • Use Case: Run a diff-scoped review after an implementation loop to produce a prioritized review report for triage or to run a full-codebase audit prior to release.

Quick Start

Run /forge-review to review changed files (diff scope) and produce a severity-tiered Markdown report saved to docs/reviews/.

Frequently Asked Questions about forge-review

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

FAQPage Schema
How do I run a multi-agent code review for algorithmic complexity and performance anti-patterns?

Run a multi-agent code review by orchestrating four specialist subagents that analyze source code in parallel for complexity issues, data structure mismatches, and performance anti-patterns, producing a deduplicated severity-tiered Markdown report.

What is the best way to automate code reviews before a release without modifying source files?

Automated code reviews can run a full-codebase audit using read-only agents that analyze paradigm violations and efficiency issues, ensuring source files remain unmodified while generating structured Markdown reports for triage.

How does severity tiering work for code review findings?

Severity tiering assigns Critical, Warning, or Suggestion levels to deduplicated code review findings, numbering each item to provide clear triage prioritization for algorithmic complexity and performance anti-patterns detected across the repository.

Can I scope a code review to only the changed files in a diff?

Yes, code reviews support diff scope to analyze only changed files within an implementation loop, as well as full-codebase scope for comprehensive audits across common programming languages and filtered file sets.

Does pipeline integration support lifecycle handoff for automated code review reports?

Pipeline integration reads loop state for lifecycle handoff and writes structured Markdown reports directly to the docs/reviews/ directory, enabling seamless downstream triage and continuous integration workflows.

What are the limitations of using multi-agent code reviews for complexity analysis?

Multi-agent code reviews are restricted to read-only analysis of source files and can only write output reports to docs/reviews/, meaning they cannot automatically apply fixes or modify code directly within the repository.