Aggregate critical feedback from multiple LLM reviewers into a findings report.
npx skills add https://github.com/raine/consult-llm --skill review-raine
Please help me install this Agent Skill. Skill: review Source: https://github.com/raine/consult-llm/tree/main/skills/review Command: npx skills add https://github.com/raine/consult-llm --skill review-raine
Collect critical, honest feedback from all LLMs on an artifact. Push back on weak arguments and surface both consensus findings and unresolved disagreements.
Provide a thorough, independent critical review of the target artifact using four reviewers and generate a findings report.
High-intent search queries and answers about installing and using this skill.
Multi-LLM critique orchestrates four independent reviewers to identify problems, gaps, and risks in artifacts like architecture docs and implementation plans. It runs a structured four-phase workflow including independent reads, cross-review, and synthesis to generate a findings report.
A multi-model code review follows a four-phase workflow: Phase 1 reads the target, Phase 2 generates independent reviews, Phase 3 cross-reviews critiques, and Phase 4 synthesizes a findings report. This enforces explicit prompts for critical evaluation.
Yes, LLM artifact reviews support configurable review sets, optional rounds, and dry-run options to tailor critique depth. These settings control the number of reviewers and the thoroughness of the synthesis process for your specific project needs.
Yes, you must load the consult-llm dependency to run multi-LLM critique. This enables the four parallel reviewers to execute independent reviews, cross-review evaluations, and final synthesis across diverse project artifacts.
Multi-LLM critique works best with artifacts such as architecture docs, implementation plans, and code reviews across diverse projects. Any technical artifact requiring rigorous critical evaluation and identification of gaps, risks, or weaknesses is suitable.
vincent-ngOrchestrate independent subagents executing per-task plans with two-stage reviews.
nand1234Revise implementation plans while preserving locked tickets in Markdown docs.
mataaninCoordinate parallel LLM code reviews and compare cross-tool agreements.
microsoftReview the current conversation with a structured six-lens critique.
