AnterisLab Guard documentation

AnterisLab Guard is the open-source client SDK for the AnterisLab policy engine. Every action an agent takes is evaluated against your policies before it executes, and a stop order actually stops the agent.

• Quick startInstall the SDK and run your first guarded action.
• ConfigurationEvery option new Guard(...) accepts.
• VerdictsWhat APPROVED, FLAGGED, BLOCKED, and PAUSED mean.
• Error handlingThe GuardError hierarchy and how to react to each.
• Kill switchLocal halt, verified state, SSE stream.
• TestingThe public mock, and testing without a subscription.
• SubscriptionFree tier, expiry, and upgrading.
• Security modelWhat the SDK protects against, and what it does not.
• API referenceThe complete public surface.

For vulnerability reporting, see SECURITY.md. For contributing, see CONTRIBUTING.md.