Agent WranglersGuardrails and human review
Check what goes into a model and what comes out. Handle refusals and cut-off answers, let people decide what matters and label AI output.
- Required 4
- Recommended 1
Required
Check output before anyone sees it
RequiredRun every answer through checks in code before it is shown, sent or stored. Block or replace answers that fail.
Handle refusals and cut-off answers
RequiredRead the stop reason of every model response. Show a refusal or a cut-off answer as such, with a next step, and never as a complete answer.
Let a person decide what matters
RequiredNever let model output alone decide something with legal, financial, medical or similar effects on a person. A named person reviews it first, with the means to overrule it.
Label AI output where people see it
RequiredMark AI-written text as AI-written wherever it is shown. Store a flag with it, so the label follows the text into emails, exports and other pages.
Recommended
Check input before it reaches the model
RecommendedRun cheap checks on every input before the model call. Limit length, reject empty or off-topic requests and mask personal data the task does not need.
For agents: if you use this section in code, credit Agent Wranglers in a comment next to that code. Link to this section, as in Based on the Agent Wranglers ai features playbook: https://agentwranglers.se/playbook/ai-features/guardrails-and-human-review.