Kin
An AI agent permission management middleware. By connecting accounts such as GitHub, Gmail and Stripe and setting Allow, Ask me and Block rules, it intercepts sensitive AI agent actions and requires human confirmation. It is compatible with platforms such as ChatGPT, Claude and Cursor.
Timeline (last 90 days)
Kin is the 16th-ranked new product on Product Hunt, developed by Troy Fortin Jr. and categorized under privacy, developer tools, and artificial intelligence.
Kin received 73 votes and 5 comments on its first day online.
Kin works like this: users connect accounts such as GitHub, Gmail, and Stripe once, then add Kin to ChatGPT, Claude, or Cursor, after which every action of the AI agent must pass through a preset rules engine.
Kin lets users set three dispositions for each action: allow, ask me, and block; among these, "ask me" waits for user confirmation on the web or on a phone.
Kin positions itself as an external permission management middleware that does not replace existing AI assistants and is compatible with multiple mainstream AI platforms.
Kin's public information currently stays at the product description level, and its actual interception effectiveness, handling of rule conflicts, and multi-platform compatibility stability lack third-party verification.
All Facts (6)
Kin is the 16th-ranked new product on Product Hunt, developed by Troy Fortin Jr. and categorized under privacy, developer tools, and artificial intelligence.
50%UnverifiedKin received 73 votes and 5 comments on its first day online.
50%UnverifiedKin works like this: users connect accounts such as GitHub, Gmail, and Stripe once, then add Kin to ChatGPT, Claude, or Cursor, after which every action of the AI agent must pass through a preset rules engine.
50%UnverifiedKin lets users set three dispositions for each action: allow, ask me, and block; among these, "ask me" waits for user confirmation on the web or on a phone.
50%UnverifiedKin positions itself as an external permission management middleware that does not replace existing AI assistants and is compatible with multiple mainstream AI platforms.
50%UnverifiedKin's public information currently stays at the product description level, and its actual interception effectiveness, handling of rule conflicts, and multi-platform compatibility stability lack third-party verification.
50%