What problem does it solve?
Generic AI-generated safety paperwork is often rejected by regulators: it uses vague tasks, defaults to PPE-only controls, has no assigned action owners, stops at "worker error" for root causes, and leaks personal data into widely shared documents. This skill eliminates those failures by enforcing structured intake to capture specific hazard and control details, applying the hierarchy of controls, mandating de-identification of sensitive data, and requiring SME review before any output is produced.
Core Features & Use Cases
- Structured Intake & Specificity Enforcement: Runs a one-question-at-a-time intake to capture the exact hazard, top event, threats, consequences, and barrier performance standards, refusing to proceed on vague requests like "a bowtie for the plant".
- Hierarchy of Controls & Barrier Validation: Automatically ranks barriers by effectiveness, flags administrative/PPE-only barrier sets without justification, and records [GAP] for missing performance standards instead of inventing unsupported data.
- Regulatory Compliance & De-identification: Applies jurisdiction-specific regulatory requirements, mandates full de-identification of personal/health data before drafting, and runs a mandatory SME review to ensure outputs are defensible for regulatory challenge.
- Use Case: A process safety manager at a chemical facility needs a bowtie diagram for LPG storage loss of containment. The skill guides them through capturing all required inputs, validates barrier performance standards, and produces a compliant, de-identified diagram ready for team and competent person review.
Quick Start
Use the bowtie-builder skill to create a compliant bowtie diagram for a named process safety hazard like LPG loss of containment, answering the structured intake questions for threats, consequences, and barrier performance standards as prompted.
