ClaudeKit

ClaudeKit is a builder-focused kit of specialized AI agents and workflows for shipping software and driving growth. It focuses on scaffolding—so you start with proven planning, coding, testing, review, and deployment flows—while keeping full ownership of how you customize and extend your system.

For engineering teams and full-stack builders, ClaudeKit Engineer (AgentKit Engineer) runs a structured, multi-stage delivery pipeline that connects specialized subagents, skills, and safety hooks into a guarded loop (Brainstorm → Plan → Cook → Review → Ship). It includes pre-built skills for engineering, research, code review, and automation, plus slash commands that cover the delivery lifecycle. The goal is to reduce “context drift” and repeated workflow setup between sessions by reusing deterministic mechanics.

For marketing and growth workflows, ClaudeKit Marketing (AgentKit Marketing) uses 32 specialized AI agents along with skills and commands aimed at research, email campaigns, programmatic SEO, and revenue growth across the funnel, supported by multiple MCP integrations for lead generation and outreach.

The kits support multiple agent runtimes via adapters, including Claude Code (plus other runtimes), and they emphasize “scaffolding, not vendor lock-in.” Pricing is available for AgentKit Engineer ($99 one-time), AgentKit Marketing ($99 one-time), the AgentKit Bundle ($149), and the AgentKit App desktop license ($49/year for 1 device or $99/year for 3 devices).

Is ClaudeKit still available, and what is it called now?

ClaudeKit has been renamed to AgentKit. The same builder-focused kits and workflows are available under the AgentKit name.

What does the engineering version (AgentKit Engineer) actually do?

It provides specialized AI subagents and skills that follow a structured pipeline from Brainstorm → Plan → Cook → Review → Ship, with safety hooks and guarded delivery stages.

Who is AgentKit best for?

It’s designed for solo founders/indie hackers and software teams that want AI-assisted development and growth workflows without rebuilding the system from scratch each time.

Does AgentKit lock me into a single platform?

No—it's positioned as scaffolding rather than vendor lock-in, letting you keep what works, change what doesn’t, and customize workflows to fit your team.

Which runtimes does it work with?

AgentKit supports multiple agent runtimes via adapters, including Claude Code and other listed harnesses/runtimes, with different support tiers (production, beta, and experimental preview).

What are the listed prices and plans?

AgentKit Engineer is $99 one-time, AgentKit Marketing is $99 one-time, and the AgentKit Bundle is $149. The desktop app license is $49/year for 1 device or $99/year for 3 devices.