self-driving-agents
Self-Driving Agents is a comprehensive agent skill library providing 179 pre-configured, self-learning AI agents organized into 13 specialized departments including Design, Engineering, Finance, Marketing, Sales, and more. Powered by Hindsight, each agent can be installed via NPX and activated in supported AI clients including Claude Chat, Claude Code, OpenClaw, NemoClaw, and Hermes. Agents range from technical specialists like Software Architects and DevOps Automators to domain experts like China Market Localization Strategists and Livestream Commerce Coaches. Developers can install built-in agents from the roster, import custom agents from local directories or GitHub repositories, or start with an empty template and build through conversation.
Key Features
Use Cases
- 01Adding specialized SEO, content marketing, or social media expertise to AI workflows
- 02Integrating domain-specific engineering agents (frontend, backend, DevOps, security) into development workflows
- 03Deploying China market specialists for Douyin, WeChat, Xiaohongshu, and Bilibili strategies
- 04Building custom game development pipelines with Unity, Unreal, Godot, or Roblox agents
- 05Creating multi-agent sales teams with discovery coaches, deal strategists, and pipeline analysts
- 06Establishing compliance and governance frameworks with identity, audit, and ZK stewardship agents
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
self-driving-agents — FAQ
What is self-driving-agents?+
Self-Driving Agents is an NPM package providing 179 pre-built, self-learning AI agent skills organized into 13 departments. Each agent is a specialized skill that can be installed and activated in AI clients like Claude Chat, Claude Code, and OpenClaw.
How do I install a built-in agent skill?+
Run 'npx @vectorize-io/self-driving-agents install <department>/<agent-name> --harness <harness-name>' (e.g., 'npx @vectorize-io/self-driving-agents install marketing/seo --harness claude'). For Claude Chat/Cowork, upload the generated skill via Customize → Skills → Upload and allowlist the API host in Settings → Capabilities.
Which AI clients work with self-driving-agents?+
Supported harnesses include Claude Chat, Claude Cowork, Claude Code, OpenClaw, NemoClaw, and Hermes. Each requires specifying the --harness flag during installation.
Can I create custom agents or use my own GitHub repos?+
Yes. Install from a local directory with 'npx @vectorize-io/self-driving-agents install ./my-agent --harness <harness>', from a GitHub repo with 'npx @vectorize-io/self-driving-agents install my-org/my-repo/path --harness <harness>', or start blank with the --empty flag.
Is self-driving-agents free to use?+
Yes, the package is open-source under the MIT license. However, you'll need access to the AI harnesses (Claude Chat, Claude Code, etc.) which may have their own subscription requirements.
What are the prerequisites or API keys needed?+
You need Node.js/NPX installed to run the installation command. Depending on your chosen harness, you'll need an account with Claude (for Claude Chat/Code/Cowork), OpenClaw, NemoClaw, or Hermes. Individual agents may require additional API keys for services they integrate with.
How do I install self-driving-agents?+
Open the source repository on GitHub and follow its README. self-driving-agents is a skill — MCP Agents Market links you directly to the official repo.
Is self-driving-agents free?+
self-driving-agents is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.