Command Your AI Fleet

The polyglot command center for AI agent orchestration. Spawn fleets, coordinate crews, triage at scale, and generate entire game worlds — across TypeScript, Python, and Rust.

Get Started View on GitHub

TypeScriptPythonRust

Packages

Languages

AI Providers

CLI

One Toolkit. Three Languages. Infinite Agents.

Agentic is the command center for AI agent orchestration. Whether you are spawning fleets of coding agents, orchestrating AI crews, triaging thousands of issues, or generating entire game worlds — you do it from one unified toolkit that speaks TypeScript, Python, and Rust.

What Can You Build?

Four capabilities, one toolkit. Each solves a real problem for teams working with AI agents at scale.

🎯

Spawn AI Agent Fleets

TypeScript

Dispatch multiple AI agents across repositories. They clone, branch, work, and open PRs — all in parallel, with automatic token routing per GitHub org.

Fleet Management Guide

🧩

Orchestrate AI Crews

Python

Define crews in YAML, run them on CrewAI, LangGraph, or Strands. One format, any framework.

Crew Docs

🔍

Triage at Scale

TypeScript

Portable Vercel AI SDK tools for issue triage, code review, and sprint planning across trackers.

Triage Docs

🎮

Generate Retro RPGs

Rust

AI-powered game generation with Bevy. Procedural worlds, characters, quests, and assets from config.

Game Generator Docs

💠

Create 3D Assets

TypeScript

Declarative 3D asset pipelines powered by Meshy AI. Define text-to-3D, rigging, and animation in JSON manifests. Built-in preview and VCR recording for CI.

Meshy Docs

The Ecosystem

Seven packages that work together or independently. Each solves a specific problem. All share the same principles: provider-agnostic, declarative config, zero hardcoded credentials.

@jbcom/agentic

TypeScript

The orchestration layer. Fleet management, token routing, sandbox execution, and multi-agent handoff protocols.

View Package

@jbcom/agentic-triage

TypeScript

Vercel AI SDK tools for issue triage, PR review, and sprint planning. Works with GitHub, Jira, Linear, and Beads.

View Package

agentic-crew

Python

Framework-agnostic crew orchestration. CrewAI, LangGraph, Strands — one YAML, any runtime.

View Package

@agentic/meshy

TypeScript

Declarative 3D asset pipelines. Text-to-3D, rigging, animation via Meshy AI with manifest state.

View Package

game-generator

Rust

AI-powered retro RPG generation with Bevy. Procedural worlds, characters, and quests.

View Package

@jbcom/agentic-providers

TypeScript

Unified provider layer for Anthropic, OpenAI, Google, Mistral, Azure, and Ollama.

View Package

@jbcom/vitest-agentic

TypeScript

Vitest utilities for testing agent workflows. Mock providers, snapshot conversations.

View Package

Quick Start

Pick your language and get running in under five minutes.

# Install the control package
npm install -g @jbcom/agentic

# Initialize configuration
agentic init

# Spawn your first AI agent
agentic fleet spawn "my-org/my-repo" \
  "Fix the failing tests" --auto-pr

# Monitor progress
agentic fleet list --running

import { getTriageTools } from '@jbcom/agentic/tools';
import { generateText } from 'ai';
import { anthropic } from '@ai-sdk/anthropic';

const result = await generateText({
  model: anthropic('claude-sonnet-4-20250514'),
  tools: getTriageTools(),
  maxSteps: 10,
  prompt: 'List critical bugs and create a triage plan',
});

console.log(result.text);

# Install with your preferred framework
pip install agentic-crew[crewai]

# Define a crew in YAML, then run it
agentic-crew run my-package analyzer \
  --input "Review this codebase for security issues"

from agentic_crew import run_crew

# Auto-detects best available framework
result = run_crew("my-package", "analyzer",
                  inputs={"code": "..."})

# Build and run the game generator
cargo run --bin vintage-game-generator

use vintage_game_generator::prelude::*;

fn main() {
    let game = QuickGenerator::new()
        .genre(Genre::Jrpg)
        .theme(Theme::Fantasy)
        .generate_and_run();
}

Who Is This For?

💻

Solo Developers

Spawn AI agents to handle the tedious work — dependency updates, code reviews, test generation — while you focus on architecture and design decisions.

🏗

Platform Teams

Standardize AI workflows across your organization. Define crews once, run them consistently. Multi-org token management handles the authentication complexity.

🎮

Game Studios

Generate retro RPG content procedurally — worlds, characters, quests, and assets. Blend classic game aesthetics with AI-driven creativity.

💠

3D Content Pipelines

Declarative Meshy-powered pipelines for text-to-3D, rigging, and animation. JSON manifests, not brittle scripts. CI-friendly with built-in VCR recording.

Design Principles

⚙

Provider Agnostic

Every package works with multiple AI providers — Anthropic, OpenAI, Google, Mistral, Ollama, Azure. Swap providers without changing code.

📄

Declarative Config

Crews in YAML. Pipelines in JSON. Fleets from CLI flags. Configuration lives in version control, not in scripts.

🔒

Security First

Token sanitization, SHA-pinned Actions, OIDC publishing, sandboxed execution. No hardcoded credentials anywhere.

🌍

Polyglot by Design

TypeScript for orchestration. Python for AI frameworks. Rust for performance. Use the right language for each job.

Explore

🚀

Getting Started

Install, configure, and spawn your first agent in under 5 minutes.

Start here

📖

Guides

Fleet management, orchestration patterns, sandbox execution, and AI triage workflows.

Read guides

🔗

Integrations

GitHub Actions, CrewAI, LangGraph, Vercel AI SDK, MCP Server — connect to your stack.

See integrations

📚

API Reference

Complete API documentation for fleet management, triage tools, and configuration.

Browse API