Tryz
ClaudecommunityYour personal Playbook for every AI agent: Claude reads your rules before it acts and checks purchases, bookings and messages with you first.
Tryz
Tryz is your personal Playbook for every AI agent: the rules, preferences and info you want every agent to follow when it acts for you, such as sizes, budget, kosher food, no flights on Shabbat, tone, and a spending ceiling. You set it once at tryz.ai and use it with Claude and other agents, instead of explaining yourself again in every chat. With Tryz connected, Claude reads your Playbook before it searches, compares or recommends, so options fit you from the first answer. Before any action that costs money, is seen by others or changes something outside the chat (a purchase, booking, food order or message), Claude checks it with Tryz. If the action breaks one of your rules, Tryz tells Claude to stop and Claude gives you a link to approve or deny the request in Tryz. A "yes" in the chat never counts as approval. Tryz never performs or approves actions itself. You choose which topics each agent may see on the consent screen, and you can change or revoke access at any time. Tools: read_playbook (read-only), check_action (records the request in your Tryz history), get_action_status (read-only), suggest (creates a suggestion that waits for your approval in Tryz).
Anthropic’s own published signals, snapshot of November 4, 2026; units undocumented — treat as ordinal.
Over time
Details
Tools(4)
- check_action
- get_action_status
- read_playbook
- suggest
Data from Claude’s public directory API, snapshotted daily by the MCP App Tracker. A listing describes what the directory publishes — not an endorsement.