arch-review

Analyze RTL architecture against microarchitecture spec and coding conventions.

43|9|Updated Feb 25, 2026
One-click install
npx skills add https://github.com/babyworm/rtl-agent-team --skill arch-review-babyworm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: arch-review
Source: https://github.com/babyworm/rtl-agent-team/tree/main/skills/arch-review
Command: npx skills add https://github.com/babyworm/rtl-agent-team --skill arch-review-babyworm

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Architecture review detects structural mismatches between the RTL implementation and the microarchitecture spec, assesses timing feasibility, and evaluates code quality, while ensuring outputs are produced without any modification to RTL code.

Core Features & Use Cases

  • Three independent reviewers (rtl-architect, timing-advisor, rtl-critic) run in parallel to evaluate structure, timing, and quality.
  • Enforces coding conventions (port prefixes i_/o_/io_, clock naming domain_clk, reset naming domain_rst_n, logic data type, instance prefix u_, and generate prefix gen_) and outputs a consolidated architecture-review.md with a Feature Coverage Checklist.
  • Outputs findings aggregated into reviews/phase-2-architecture/architecture-review.md, including a prepared standard review Markdown format.

Quick Start

Run the three read-only agents to generate architecture review findings and store results in reviews/phase-2-architecture/architecture-review.md.

Frequently Asked Questions about arch-review

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

FAQPage Schema
How do I perform a read-only RTL architecture review before tape-out?

Perform a read-only RTL architecture review by running three parallel agents to evaluate structure, timing, and maintainability against the microarchitecture spec without modifying any RTL code, generating a consolidated architecture-review.md.

What coding conventions are enforced during an RTL architecture sign-off?

RTL architecture sign-off enforces coding conventions including i_/o_/io_ port prefixes, domain_clk clock naming, domain_rst_n reset naming, logic data types, u_ instance prefixes, and gen_ generate prefixes to ensure structural consistency.

When do I need to run a pre-tape-out architecture review for my RTL design?

Run a pre-tape-out architecture review after completing RTL implementation and before tape-out gates to detect structural mismatches with the microarchitecture spec, assess timing feasibility, and evaluate code quality.

How does the three-reviewer parallel approach work for RTL architecture analysis?

The three-reviewer parallel approach uses rtl-architect, timing-advisor, and rtl-critic agents running simultaneously to independently evaluate RTL structure, timing feasibility, and code quality, aggregating findings into a single review document.

What artifacts are generated by an RTL architecture review for sign-off?

RTL architecture review generates a consolidated architecture-review.md file stored in reviews/phase-2-architecture/, featuring a Feature Coverage Checklist and a prepared standard review Markdown format for sign-off validation.

Can I validate microarchitecture spec consistency without modifying my RTL code?

Validate microarchitecture spec consistency without modifying RTL code by running read-only architecture reviews that analyze structural mismatches, timing feasibility, and code quality, producing a consolidated review document for sign-off.