clawdirect-dev
Build agent-facing web experiences with ATXP-based authentication, following the ClawDirect pattern. Use this skill when building websites that AI agents interact with via MCP tools, implementing cookie-based agent auth, or creating agent skills for web apps. Provides templates using @longrun/turtle, Express, SQLite, and ATXP.
Security Assessment
Detected risks:
About clawdirect-dev
The `clawdirect-dev` skill enables developers to build agent-facing web experiences with ATXP-based authentication, addressing the need for secure and programmatically accessible web applications for AI agents. It provides a structured approach to implementing cookie-based authentication for agents, bridging interactions between browser automation and MCP tool endpoints. By leveraging the ClawDirect pattern, this skill ensures agents can authenticate, navigate, and interact with web interfaces efficiently while maintaining security and identity verification. It eliminates the complexity of handling direct agent authentication in web environments and streamlines the process of exposing web actions to AI agents.
Key features include integration with ATXP for agent identity and optional payment handling, MCP tool support for programmatic actions, and a reference architecture using Express, SQLite, and @longrun/turtle. The skill provides templates and step-by-step guidance for setting up the MCP server, implementing cookie-based authentication, and publishing agent skills. It also addresses practical challenges such as handling HTTP-only cookies for browser-based agents by using query string token passing, ensuring smooth and secure agent access.
This skill is particularly useful for developers creating AI agent-enabled websites, SaaS platforms, or web applications that need controlled agent interactions. Target users include full-stack developers familiar with Node.js, TypeScript, and web frameworks like Express, as well as teams building AI-driven automation workflows. Use cases include creating agent-accessible dashboards, automating interactions for AI agents, and offering premium web actions that require authenticated access or payment tracking, all while maintaining a robust and secure architecture.
FAQ
How do agents authenticate when using this skill?
Agents authenticate using ATXP-based mechanisms. They receive an authentication cookie via MCP tools, which is then used in their browser interactions with the site.
Which technologies are required to use this skill?
The skill requires Node.js with Express for the web server, @longrun/turtle for MCP server integration, SQLite for storing auth cookies, and optionally TypeScript for type safety.
Can agents directly set HTTP-only cookies in their browsers?
No, HTTP-only cookies cannot be set directly by agent browsers. The recommended approach is to pass the cookie value in the query string, have the server set the cookie, and redirect to a clean URL.
Is payment handling mandatory for agent interactions?
No, payment handling through ATXP is optional. It can be used for charging premium actions but is not required for basic authentication and agent interactions.
Who is the intended audience for this skill?
The skill targets developers and teams building AI agent-enabled web applications who need secure authentication, programmatic access via MCP, and integration with web frontends.
Install clawdirect-dev
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
napoleond/clawdirect