Coordinate parallel multi-model reviews to identify and prioritize code defects.
npx skills add https://github.com/raine/consult-llm --skill review-panel-raine
Please help me install this Agent Skill. Skill: review-panel Source: https://github.com/raine/consult-llm/tree/main/skills/review-panel Command: npx skills add https://github.com/raine/consult-llm --skill review-panel-raine
Standalone multi-model code review of an existing diff. Multiple LLMs review in parallel; agent deduplicates, prioritizes by severity/confidence, and optionally applies localized fixes.
Load the consult-llm skill and run a diff review to have multiple models assess the changes in parallel.
High-intent search queries and answers about installing and using this skill.
To run a multi-model code review, load the consult-llm skill and apply it to your diff. Multiple LLMs will review the changes in parallel, and the agent synthesizes the findings into a prioritized checklist for triage.
The best way to triage defects is to use a coordinated multi-model review process. This approach identifies and surfaces defects in code changes by deduplicating findings and prioritizing them by severity and confidence.
Yes, you can apply localized fixes automatically. The agent optionally applies unambiguous fixes after synthesizing the prioritized findings from the independent parallel reviews of your software diff.
Multi-model code review deduplicates and prioritizes findings by having an agent synthesize the parallel outputs. The agent applies a consistent rubric to produce unified priorities based on severity and confidence for triage.
You need to provide your software diffs, changed files, and baselines across repositories. The review process uses these inputs to prompt the multiple models with a consistent rubric and output structured findings.
You should avoid parallel multi-model reviews when your code changes are ambiguous and require localized fixes, as the automated remediation is strictly optional and only applies where fixes are unambiguous.
mkusakaTrace changed functions to callers and validate call-site contracts from diff files.
ZateGenerate structured security audit reports with findings, severity, and ASVS mappings.
raineQuery external LLMs for second opinions on files with model flags.
ArogyaReddyCoordinate parallel code reviews and consolidate findings into a unified report.
