Privacy should be clear before a user connects an AI agent.
AgentShield exists to help teams inspect risky AI behavior. This page explains what the app needs to store, why it stores it, and what users should avoid sending.
Last updated: May 28, 2026
We keep the product understandable
The dashboard shows what was blocked, why it was blocked, and when it happened. We avoid hiding safety decisions behind vague labels.
We store only what the product needs
AgentShield stores account, rule, API key metadata, and interceptor logs so you can debug and audit your AI agent behavior.
API keys are sensitive
Treat generated keys like passwords. If a key is exposed, revoke it from the dashboard and create a fresh one.
What AgentShield stores
Account data
Used for login, ownership, and dashboard access.
API keys
Used to authenticate requests sent to the interceptor.
Rules
Used to decide whether an input or output should be blocked.
Logs
Used to show clean/blocked events, matched rules, and timestamps.
Avoid sending secrets in test prompts
Use realistic examples, but do not paste live credentials, private customer records, or production secrets while testing. If you need help with a sensitive setup, contact us first.
Privacy questions?
Ask before connecting a production agent or sensitive workflow.