What problem does it solve?
Generic AI generates vague, regulator-rejectable safety paperwork for toxic release scenarios, often inventing unsupported dispersion distances, leaking sensitive personnel data, and missing critical source term inputs. This skill eliminates those failures by enforcing structured, specificity-first intake, mandatory de-identification, and strict assistive boundaries that prevent autonomous quantitative modelling.
Core Features & Use Cases
- Structured Intake: Forces capture of exact substance, inventory, release mode, receptors, and target study via a branched multi-step questionnaire, refusing to proceed on vague requests like "a toxic gas leak".
- No Invented Modelling: Never runs PHAST/ALOHA or invents dispersion distances/concentrations; all unmodelled values are flagged [GAP] and handed to competent modellers.
- Built-in Compliance Guardrails: Mandatory de-identification of all personal and health data, hierarchy of controls application, and mandatory SME review before any output is delivered.
- Use Case: A process safety engineer at a chemical plant uses this skill to structure a chlorine release scenario input for their team's bowtie study, ensuring all required source term and receptor data is captured without risking non-compliant fabricated dispersion figures.
Quick Start
Use the toxic-release-dispersion-scenario skill to frame a toxic release scenario from your site's substance inventory, release details, and receptor information for your process safety bowtie or LOPA study.
