blackbox

Automates multi-model coding tasks with built-in judge evaluation via Blackbox CLI.

Updated Apr 1, 2026
One-click install
npx skills add https://github.com/founderphantom/zola-agent --skill blackbox-founderphantom
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: blackbox
Source: https://github.com/founderphantom/zola-agent/tree/main/optional-skills/autonomous-ai-agents/blackbox
Command: npx skills add https://github.com/founderphantom/zola-agent --skill blackbox-founderphantom

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Blackbox AI CLI delegates coding tasks to a multi-model agent that runs tasks through multiple models and uses a built-in judge to select the best result.

Core Features & Use Cases

  • Multi-model evaluation and selection: tasks are processed by several models (e.g., Claude, Codex, Gemini) and the best implementation is chosen.
  • Interactive and background tasks: supports one-shot sessions, interactive prompts, and background processing with progress monitoring.
  • Parallel work and PR reviews: spawn isolated sub-tasks, clone repos, review changes, and consolidate results.
  • Session management and prompts: configurable providers, models, and token usage with structured feedback.

Quick Start

Prompt Blackbox to implement a coding task and compare outputs across models to select the best solution.

Frequently Asked Questions about blackbox

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a coding task across multiple AI models to compare outputs?

To compare coding task outputs across multiple AI models, you can delegate your prompt to a multi-model CLI agent. The system processes the task through several models and uses a built-in judge to evaluate and select the best implementation for your code.

Can I use a coding agent to review pull requests and consolidate changes?

Yes, you can use a multi-model coding agent to review pull requests. It spawns isolated sub-tasks, clones repositories, reviews changes, and consolidates results to streamline your software development workflow and parallel work.

How does a multi-model AI CLI select the best code implementation?

A multi-model AI CLI selects the best code implementation by utilizing a built-in judge workflow. After processing the coding task through various configured models, the judge evaluates the results to ensure quality across one-shot and background executions.

Does the multi-model coding agent support interactive prompts and background processing?

Yes, the multi-model coding agent supports interactive prompts, one-shot sessions, and background processing. It provides progress monitoring so you can track iterative refinements and complex coding tasks as they run.

What's the best way to automate rapid prototyping with a multi-model coding agent?

The best way to automate rapid prototyping is to delegate coding tasks to a multi-model agent via CLI. It handles model configuration and token usage, allowing you to quickly generate, evaluate, and refine code implementations.

Do I need to configure providers and models before using a multi-model CLI?

Yes, you need to configure providers, models, and token usage before using the multi-model CLI. The system implements prerequisites and model configuration to ensure structured feedback and quality results across executions.