What problem does it solve?
Generic AI-generated HAZID outputs are often unstructured, miss critical external and environmental hazards, fabricate risk ranks or controls to fill gaps, or leak sensitive participant and injury data. This skill eliminates those failures by enforcing a structured, team-led intake process, mandatory de-identification, strict no-fabrication rules, and hierarchy of controls validation to produce defensible, regulator-ready hazard registers.
Core Features & Use Cases
- Structured Multi-Step Intake: Guides users through a one-question-at-a-time elicitation process to bound the HAZID scope (named installation/life-cycle stage), capture all required hazard categories (process, mechanical, electrical, external/natural, environmental, utilities loss, neighbouring knock-on), identify receptors for external/environmental hazards, and confirm team competencies.
- Assistive Team Recording: Records only the hazards, causes, consequences, and controls the team supplies, never inventing unraised hazards or fabricating risk ranks — uses
[GAP] for missing or unaddressed items to maintain full transparency.
- Compliance & Validation Built-In: Applies hierarchy of controls ranking to all recommendations, enforces mandatory de-identification of sensitive data, runs pre-output quality checks, and includes a mandatory SME review gate to ensure outputs meet regulatory and industry standards.
- Use Case: A process safety engineer preparing for a new tank farm FEED stage HAZID can use this skill to structure the workshop, capture all relevant hazards (including external flood and seismic risks), rank controls, and generate a de-identified, compliant hazard register ready for competent person review.
Quick Start
Use the hazid-facilitator skill to structure a HAZID workshop for your new chemical process installation at the FEED stage and generate a de-identified hazard register with risk-ranked findings and control recommendations.
