agentix.guards.base

base

Guard primitives: the uniform checkpoint the loop runs around every tool call.

A :class:Guard exposes two optional hooks:

• before_call — inspect a pending :class:~agentix.types.ToolCall and return a :class:Decision (allow / deny / confirm).
• after_output — transform a tool's output text before it re-enters the model's context (e.g. neutralize injection, mark as untrusted data).

A :class:GuardPipeline runs an ordered list of guards. before_call stops at the first deny; any confirm along the way means the loop must get a human "yes" before executing. This replaces the reference's hard-coded if ladder with composable, swappable objects.

Decision dataclass

Decision(type: DecisionType, reason: str = '')

A guard's verdict on a pending tool call.

GuardContext dataclass

GuardContext(policy: AgentPolicy)

Read-only context handed to every guard for a given call.

Guard

Base guard. Subclass and override the hooks you need; defaults are no-ops (allow / pass-through), so a guard only implements what it cares about.

Three checkpoints, covering both boundaries: * before_call — a pending tool call (ingress to a tool). * after_output — a tool's result re-entering context (egress from a tool). * on_answer — the model's final answer leaving for the user (egress to the user). Use it for redaction / DLP on what the user sees.

GuardPipeline

GuardPipeline(guards: Sequence[Guard] = ())

Runs an ordered list of guards as a single checkpoint.