Trustra Guardrails

Stop describing your policy. Enforce it.

A guardrail written in a policy constrains nothing. Trustra Guardrails screens inputs, checks outputs, bounds what agents may do, and caps what they may spend, then records every allow, block, and override on the same tamper-evident chain as your audit evidence.

What it does

Four layers of enforcement, one record.

Guardrails act at the point of use, before harm or spend occurs, and leave evidence that they ran.

01

Input screening

Detect prompt injection and jailbreak attempts, screen personal data before it reaches a third-party model, and reject requests outside the sanctioned purpose.

02

Output checks

Verify answers are grounded in approved sources, filter unsafe content, catch personal data and system prompt leakage, and enforce AI disclosure.

03

Agent action limits

Allowlist the tools an agent may call, scope credentials to least privilege, bound blast radius, and require confirmation before irreversible actions.

04

Cost and consumption caps

Token and request quotas per user or tenant, context and output caps, retry and loop limits, so a runaway agent cannot exhaust your budget or your capacity.

05

Defined behavior at the limit

Every limit has a decided outcome: degrade to a cheaper model, queue the request, or refuse with a clear message. No silent failures.

06

Override trail

When a human bypasses a guardrail, the approver, the justification, and the outcome are recorded. Override patterns become a governance signal.

Why it matters

A control you cannot evidence is a control that may not be running.

Most organizations can show a guardrail exists in a document. Far fewer can prove it was active on a specific date. Trustra closes that gap by treating enforcement and evidence as the same act.

Prompt injectionJailbreaksPII leakageUngrounded answersToxic outputDisclosure gapsRogue tool callsRunaway spend
1,204
CHECKS / 24H
18
BLOCKED
3
OVERRIDES
100%
SEALED
input: injection12 blocked
output: grounding0.96 pass
action: tool scope2 denied
cost: quota1 throttled
chainverified
How it connects

Enforce in line, or observe first.

Start in monitor mode to see what would have been blocked, then switch to enforcement once thresholds are calibrated. Either way, every decision lands on the evidence chain.

API gateway

Enforce in line

Route model calls through the Trustra gateway and guardrails apply before the request reaches the provider.

Two-line SDK

Enforce in your app

Wrap your calls and apply the same policy locally. pip install trustra

Monitor mode

Measure before you block

Run every check without enforcing, and review what would have been blocked before you turn it on.

Make your guardrails provable.

Start free and see which checks fire on your own traffic, in monitor mode, before enforcing anything.

The platform

Explore the rest of the platform