Does this skill require specific file formats?

Yes, it is designed to work with OpenSpec plan artifacts, including proposal.md, design.md, and tasks.md files located in the project's changes directory.

Can I use this for security audits?

Yes, the skill includes a dedicated security review step that evaluates plans for input validation, authentication/authorization gaps, and OWASP top-10 vulnerabilities.

What happens if the different AI vendors disagree?

The consensus synthesizer identifies disagreements and flags them for escalation to a human developer, ensuring that conflicting architectural or security opinions are resolved manually.

What is a parallel review plan?

It is a specialized workflow where an implementation plan is independently evaluated by one or more AI models to identify errors, security risks, or architectural inconsistencies before a single line of code is written.

How does multi-vendor consensus work?

The skill dispatches the review prompt to various LLM providers (e.g., Claude, Gemini, GPT) configured in your environment. It then synthesizes their findings into a single report, highlighting points of agreement and areas requiring human intervention.

Parallel Review Plan

Name: Parallel Review Plan
Author: jankneumann

byjankneumann

•

セキュリティとテスト

Conducts independent, multi-vendor architectural and security reviews of technical plans to ensure robust implementation.

The Parallel Review Plan skill automates the validation of technical proposals, architectural designs, and task breakdowns before code implementation begins. By analyzing OpenSpec artifacts against a rigorous checklist—covering security, contract consistency, and architectural alignment—it identifies potential issues early in the development lifecycle. A standout feature of this skill is its ability to dispatch reviews across multiple LLM vendors to synthesize a consensus, significantly reducing AI hallucinations and ensuring a peer-reviewed foundation for complex agentic coding tasks.

主な機能

012 GitHub stars

02Analysis of work package dependency graphs to ensure implementation feasibility

03Multi-vendor consensus synthesis to reduce AI bias and improve review accuracy

04Structured JSON output conforming to open-source review-findings schemas

05Automated architectural alignment and security vulnerability checklists

06Validation of contract consistency between OpenAPI schemas and specifications

ユースケース

01Validating complex feature proposals before handing them off to implementation agents

02Performing cross-vendor architectural audits for high-stakes enterprise projects

03Automating security and compliance checks during the planning phase of a development sprint

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jankneumann/agentic-coding-tools parallel-review-plan

For use in Claude.ai and ChatGPT

主な機能

012 GitHub stars

02Analysis of work package dependency graphs to ensure implementation feasibility

03Multi-vendor consensus synthesis to reduce AI bias and improve review accuracy

04Structured JSON output conforming to open-source review-findings schemas

05Automated architectural alignment and security vulnerability checklists

06Validation of contract consistency between OpenAPI schemas and specifications

ユースケース

01Validating complex feature proposals before handing them off to implementation agents

02Performing cross-vendor architectural audits for high-stakes enterprise projects

03Automating security and compliance checks during the planning phase of a development sprint

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jankneumann/agentic-coding-tools parallel-review-plan

For use in Claude.ai and ChatGPT