llm-external-review

An external LLM code review avoids AI bias by invoking independent models via CLI to analyze code and architecture. This keeps the primary assistant from reviewing its own logic, providing diverse viewpoints for critical refactors or design decisions.

External architecture reviews work by preparing a context file and invoking the review.py script with a chosen model CLI. The script returns the external model's feedback for synthesis, ensuring an independent perspective on design plans.

Yes, you can use external LLM models to obtain second opinions on software design decisions. The workflow invokes external models via CLI to review architecture and code, returning independent feedback to brainstorm alternatives and validate approaches.

The best way to run an independent code review without built-in bias is to invoke external LLM models through a CLI script. This delegates the actual review process to an external model, returning fresh feedback for synthesis.

Yes, you need an appropriate model CLI installed to perform external code reviews. The workflow requires running the review.py script with a chosen model CLI to invoke the external LLM and return its feedback.

You should avoid external architecture reviews when an independent perspective is unnecessary or when context files cannot be safely shared with external model CLIs. It is intended for critical software engineering tasks where avoiding bias and obtaining diverse viewpoints are valuable.