# HALO Engine

An AI Orchestration Risk Scan evaluating submitted responses across orchestration, state, authority, permissions, guardrails, observability, integration, and economics.

Created by David Sidney Williams through ThinkHALO. ThinkHALO publishes this current system; David’s canonical identity is on his personal domain.

## Authoritative resources

- [HALO Engine](https://engine.thinkhalo.ai/)
- [David Sidney Williams](https://davidsidneywilliams.com/)
- [ThinkHALO](https://thinkhalo.ai/)
- [Current practice description](https://thinkhalo.ai/ai)

The scan evaluates submitted responses. Rules calculate scores; AI interpretation does not change those scores. The application controls the scan experience and its results.

## Knowledge

- [AI orchestration knowledge](https://engine.thinkhalo.ai/knowledge): practical explanations of the conditions the diagnostic evaluates.
- [Authority in AI workflows](https://engine.thinkhalo.ai/knowledge/authority): assign an authoritative system for each data domain and mutation type; resolve conflicting updates. Authority is distinct from Permissions (allowed access and changes) and Guardrails (validation, approval, and stopping actions).