Coordinate cross-model reviews of architectural, code, or design decisions using Gemini 3.1 Pro and GPT-5.2.
npx skills add https://github.com/markusstrasser/skills --skill model-review
Please help me install this Agent Skill. Skill: model-review Source: https://github.com/markusstrasser/skills/tree/main/model-review Command: npx skills add https://github.com/markusstrasser/skills --skill model-review
Cross-model reviews automate adversarial critique of architectural, code, or design decisions by orchestrating multiple AI models to surface blind spots, inconsistencies, and improvement opportunities.
Initiate a cross-model review by invoking the model-review skill on your project topic or decision.
High-intent search queries and answers about installing and using this skill.
To run a cross-model architecture review, invoke the skill on your project decision to coordinate independent critiques using Gemini 3.1 Pro and GPT-5.2, which outputs issue extraction, disposition, and synthesis.
Cross-model critique automates adversarial code analysis by orchestrating multiple AI models to surface blind spots and inconsistencies, providing diverse perspectives that strengthen architectural and design decisions.
Yes, cross-model design critiques require llmx, Gemini 3.1 Pro, and GPT-5.2 to independently evaluate decisions and synthesize recommendations for your software projects.
Yes, cross-model code analysis supports constitutional anchoring via CONSTITUTION.md to guide the independent evaluation and synthesis process for architectural and design trade-offs.
The best way to analyze design trade-offs is using cross-model review in brainstorming mode, which synthesizes independent evaluations from Gemini 3.1 Pro and GPT-5.2 into structured recommendations.
Automated cross-model reviews should be avoided for simple projects not requiring independent critique or when lacking the required Gemini 3.1 Pro and GPT-5.2 dependencies.
wtthorntonCoordinate YAML-defined workflows and gate decisions with scoring data.
AreteDriverAnalyze code for bugs, security issues, and missing tests.
markusstrasserRoute prompts to optimal frontier models usingtask-specific guidance and checklists.
peelCoordinates AI model debates to synthesize architectural design recommendations.
