How to Add Human Approval to an AI Workflow
Design approval queues that prevent premature emails, risky changes and silent automation errors.
Read the guide →Guides by topic
Category: Governance & Safety
Approval, privacy, security and error recovery.
Privacy Checklist for US, UK and EU AI Workflows
A jurisdiction-aware data-minimization checklist for customer-facing AI automation.
Read the guide →How to Evaluate an AI Agent With a Small, Honest Test Set
A repeatable test plan with ordinary, ambiguous, risky and adversarial cases for small workflows.
Read the guide →AI Workflow Failure Recovery: Retries, Rollbacks and Kill Switches
A practical incident response plan for failed integrations, duplicate actions and unintended AI output.
Read the guide →Prompt Injection in Connected AI Workflows: A Small Business Defence
Stop untrusted emails, pages and documents from becoming instructions to connected agents.
Read the guide →What Not to Automate With AI: A Risk-First Decision Matrix
A practical list of high-impact actions that need human accountability, verified data and careful controls.
Read the guide →