# Mandorka Research

> The validation program, evidence limits and research discipline behind WorkProof.

Mandorka is a candidate-owned, publicly verifiable WorkProof network for proving real AI-assisted work capability through bounded tasks, deterministic server-side scoring and canonical proof records.

Canonical HTML: https://mandorka.com/research

## Key facts

- Predictive-validity claims remain explicitly limited until sufficient real-world outcome data exists.
- Methodology documents what each score dimension measures, what the evidence can support, and what it cannot support.
- Completed evidence can be published as a canonical, publicly verifiable WorkProof record with integrity metadata.

## Current assessment catalog

- **20 assessment tracks**
- **160 questions across the current banks**

## Related Mandorka authority

- [WorkProof](https://mandorka.com/workproof/index.md): The protocol for candidate-owned, versioned, publicly verifiable evidence of real AI-assisted work.
- [WorkProof Standard](https://mandorka.com/standard/index.md): The public specification for interoperable Mandorka WorkProof evidence.
- [Methodology](https://mandorka.com/methodology/index.md): How Mandorka structures assessments, scoring dimensions, evidence boundaries and interpretation.
- [Research](https://mandorka.com/research/index.md): The validation program, evidence limits and research discipline behind WorkProof.
- [Verification Center](https://mandorka.com/verify/index.md): The public entry point for checking canonical Mandorka WorkProof records.

## Limitations

Mandorka does not claim universal employment thresholds, identity verification or predictive validity until supported by sufficient real-world validation data.