shatter-and-recalibrate
Shatter compromised calibration and rebuild from ground truth.
When & Why to Use This Skill
The shatter-and-recalibrate skill is a high-integrity diagnostic and corrective utility designed to maintain the accuracy of AI reasoning and data processing. It identifies 'compromised calibration'—instances where logic, parameters, or data have drifted into error or bias—and systematically rebuilds the operational context from 'ground truth' (verified, primary data). This ensures that automated systems remain reliable, objective, and aligned with factual evidence.
Use Cases
- AI Hallucination Correction: Detecting when an agent's internal reasoning has become 'compromised' by false patterns and forcing a reset based on verified ground-truth documentation.
- Data Integrity Audits: Periodically 'shattering' existing data assumptions in a project to rebuild the analysis from scratch using only verified primary sources, preventing the propagation of errors.
- Systemic Bias Mitigation: Re-aligning automated decision-making workflows that have developed systematic biases over time, ensuring outputs match actual performance benchmarks and ground-truth reality.
| name | shatter-and-recalibrate |
|---|---|
| description | Shatter compromised calibration and rebuild from ground truth. |
Instructions
- Initialize shatter-and-recalibrate operational context
- Execute primary protocol actions
- Validate results and generate output
Examples
- "Execute shatter-and-recalibrate protocol"
- "Run shatter and recalibrate analysis"