What problem does it solve?
Generic AI-generated mining incident documentation fails regulatory review: it uses vague task descriptions, defaults to PPE-only controls, assigns blame to individual miners instead of systemic factors, invents unverified DGMS form numbers, and leaks sensitive personal and health data in widely shared reports. This Skill eliminates those failures by enforcing de-identification-first processing, structured fact elicitation, and strict adherence to verified regulatory anchors.
Core Features & Use Cases
- De-identification-First Workflow: Scrubs all personal/health identifiers, exact mine locations, and suppresses small (<5) injury/fatality cells before any analysis, with a separate re-identification key kept access-controlled.
- ICAM-Led Root Cause Analysis: Runs ICAM (default) or other selected RCA methods to identify systemic and organisational contributing factors, never stopping at "miner error", with every cause tied to a numbered evidence item.
- DGMS Reportability Verification: Resolves the mine's DGMS region/zone first, cites only verified 24-hour accident notice and Form J anchors, and marks all unverified form IDs as [GAP] instead of inventing them.
- Hierarchy-of-Controls CAPA: Drafts corrective actions ranked by the hierarchy of controls, each tracing to a named root cause with a designated owner and due date.
- Use Case: For a roof fall at an underground coal mine in the Dhanbad DGMS zone, use this Skill to investigate the incident, run ICAM RCA to identify organisational contributing factors, confirm the 24-hour notice and Form J reporting requirements, and draft a compliant corrective action plan with assigned owners.
Quick Start
Use the mine-incident-investigation skill to investigate your mining incident by providing event details, DGMS region/zone, and available evidence through the structured intake to receive a de-identified, DGMS-compliant report with root cause analysis and corrective actions.
