llm-assist

What problem does it solve?

This Skill helps you get independent help from another LLM when you need a stronger answer than a single model can provide. It is designed for debugging, code review, planning, verification, root cause analysis, rescue after repeated failures, and direct code questions.

Core Features & Use Cases

• Cross-model collaboration: Runs Claude, Codex, or OpenCode as an external thinking partner so you can compare perspectives.
• Safe prompt assembly: Builds prompt files with project instructions, protected quoting, and monitored execution to avoid brittle ad hoc CLI calls.
• Structured workflows: Supports review, debug, plan, verify, rca, rescue, and ask modes with clear provider selection and output synthesis.
• Use case: Ask it to review a tricky diff, confirm a fix, or investigate a bug when you want a second opinion before changing code.

Quick Start

Ask the llm-assist skill to run the complementary external LLM on your current code problem and return a synthesized answer.