blackbox

Delegate coding tasks to a multi-model AI agent via Blackbox CLI.

1|1|Updated May 25, 2026
One-click install
npx skills add https://github.com/aayushsoam/clawbot-agent --skill blackbox-aayushsoam
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: blackbox
Source: https://github.com/aayushsoam/clawbot-agent/tree/main/optional-skills/autonomous-ai-agents/blackbox
Command: npx skills add https://github.com/aayushsoam/clawbot-agent --skill blackbox-aayushsoam

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires blackbox-cli, and includes scripts (resource) components.

What problem does it solve?

This Skill simplifies coding tasks by delegating them to the Blackbox AI CLI agent, which handles multiple LLMs and selects the best implementation.

Core Features & Use Cases

  • Multi-Model Execution: Runs tasks through multiple LLMs (Claude, Codex, Gemini, Blackbox Pro) and picks the best result.
  • Integration with Blackbox CLI: Uses the Blackbox CLI for interactive sessions, non-interactive one-shots, checkpointing, MCP, and vision model switching.
  • Use Case: Need to refactor an auth module? Simply provide the task and let Blackbox handle the rest, leveraging its multi-model capabilities and built-in judge.

Quick Start

Delegate the task of adding JWT authentication with refresh tokens to the Express API using the Blackbox AI CLI: blackbox --prompt 'Add JWT authentication with refresh tokens to the Express API'.

Frequently Asked Questions about blackbox

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

FAQPage Schema
How do I run coding tasks through multiple LLMs and pick the best implementation?

Multi-model code execution runs your coding task through Claude, Codex, Gemini, and Blackbox Pro, using a built-in judge to automatically evaluate and select the best implementation.

What is the best way to refactor an auth module using an AI agent?

Refactoring an auth module is handled by delegating the task to the Blackbox AI CLI agent, which processes the prompt and leverages multiple language models to generate the refactored code.

Do I need the blackbox-cli dependency to delegate coding tasks?

Yes, the blackbox-cli dependency is required, as it provides the underlying interactive sessions, non-interactive one-shots, checkpointing, MCP, and vision model switching used for task execution.

How does a built-in judge work when executing code through multiple language models?

The built-in judge evaluates the code implementations generated by the multiple LLMs during execution, automatically selecting the single best result to return for your coding task.

Can I add JWT authentication with refresh tokens to an Express API using a multi-model AI agent?

Yes, you can automate adding JWT authentication by providing a prompt to the AI agent, which executes the task across multiple models and selects the best Express API implementation.