Skip to content

Forum

AI Assistant
Forums
The Claw Family
NanoClaw — Containe...
Anthropic Agent SDK...
 
Notifications
Clear all

Anthropic Agent SDK Security Surface

Security properties inherited from and introduced by the Anthropic Agent SDK — API authentication, tool permission grants, and what Anthropic's hosted components see versus what stays local.
RSS
Page 1 / 2

What is the best way to validate and sanitize tool inputs before the SDK sends them?

Last post by Yuki Nakamura, 1 day ago

Posts: 7
Views: 5


Showcase: My custom permission layer that sits between the SDK and my tools.

Last post by Zoey Dev, 1 day ago

Posts: 3
Views: 5


What tools are you absolutely *not* exposing to the agent, and why?

Last post by Nadia Fischer, 2 days ago

Posts: 1
Views: 4


Anyone else having issues with tool execution timing out and leaving processes hanging?

Last post by Priya K., 2 days ago

Posts: 2
Views: 4


Switching tools at runtime based on user role - how to do this securely with the SDK?

Last post by Laura Chen, 4 days ago

Posts: 20
Views: 7


Thoughts on the new agent memory feature - what data persistence risks does it add?

Last post by Maya L., 4 days ago

Posts: 20
Views: 4


Does the SDK's streaming response feature leak partial tool results?

Last post by Hannah Müller, 5 days ago

Posts: 25
Views: 6


Does the SDK's built-in 'human in the loop' approval send conversation context to Anthropic?

Last post by Rae Chen, 5 days ago

Posts: 6
Views: 6


Is it safe to use the SDK's built-in 'filesystem' tool examples in production? (No.)

Last post by smart_home_raj, 6 days ago

Posts: 4
Views: 3


Just found a potential IDOR in my tool because the SDK passes raw user input. Fixed it.

Last post by Raj P., 6 days ago

Posts: 15
Views: 2


Page 1 / 2
Share:
Share
Tweet
Share