Valet.dev provides a secure platform for sharing AI-generated artifacts through private, company-controlled URLs. It works with major AI agents like Claude Code, Cowork, ChatGPT, and Codex, ensuring that all shared work remains accessible only to authorized team members. The platform also supports live data integration through MCP servers, allowing published pages to display real-time information.
Funding
Funding not disclosed
Founders
Product
Problem
AI-generated work is often scattered across various agents, public chat links, web servers, or localhost environments, creating security risks and making it difficult for teams to find and access shared artifacts. Public links expose sensitive work, while localhost limits accessibility, and there is no centralized, private home for all AI-created content.
Solution
Valet.dev offers a secure, centralized platform where AI agents can publish artifacts to a company-specific domain, generating URLs that only authorized team members can open. The platform integrates directly with popular AI agents like Claude Code, Cowork, ChatGPT, and Codex, as well as any MCP client, allowing for seamless publishing without additional setup. Valet ensures that all shared work adheres to a team's design system, maintains privacy by keeping content off the public internet, and supports live updates by connecting to MCP servers for real-time data. It also provides a searchable directory for all artifacts, making it easy to find and iterate on work, with features like commenting and tagging for team collaboration.
Target Audience
Primary customers are software development teams and organizations that use AI agents for content creation and need a secure, centralized way to share and manage AI-generated artifacts.
Features
- Integration with major AI agents including Claude Code, Cowork, ChatGPT, Codex, and any MCP client
- Company-specific domain for all published artifacts, ensuring private, access-controlled URLs
- Design system enforcement so all shared artifacts follow team-specific fonts, colors, and layout rules
- Live page support via MCP server connections, allowing artifacts to display real-time data from sources like Stripe, Snowflake, and Linear
- Centralized, searchable directory for all AI-generated work
- Commenting and tagging capabilities for team collaboration directly on the published page
- Access revocation when a team member leaves, ensuring security is maintained