Credential Vault
A secret storage and retrieval mechanism located outside the AI agent's conversational context, in which a tool process retrieves credentials on behalf of the model, ensuring that raw secret values do not enter text the model can read or write.
Timeline (last 90 days)
A viable credential security pattern is to use a credential store or vault located outside the conversation: the model initiates an action request, the tool retrieves the secret and uses it directly, and the model sees a reference or the result rather than the value itself.
The key criterion for verifying whether a credential vault is effective is this: when an agent uses a secret, the value does not appear in the conversation record; if it does, the vault is merely decoration.
All Facts (2)
A viable credential security pattern is to use a credential store or vault located outside the conversation: the model initiates an action request, the tool retrieves the secret and uses it directly, and the model sees a reference or the result rather than the value itself.
50%UnverifiedThe key criterion for verifying whether a credential vault is effective is this: when an agent uses a secret, the value does not appear in the conversation record; if it does, the vault is merely decoration.
50%