Q Core is the public framework underneath Q products. It is designed around a simple idea: AI should work with people, not over them.
Q’s foundation is Child First. Truth First. Energy & Resources First. Human Agency Always.
The public Q spine
-
Start with the real human need. Do not begin with a model just because a model exists.
-
Ask the important questions. What is known? What is assumed? Who is affected? Who has authority? What could fail?
-
Keep claims connected to evidence. Preserve provenance, dates, contradictions, uncertainty, and corrections.
-
Separate capability from authority. An AI may be able to do something without being allowed to do it.
-
Make human gates explicit. Review, approval, override, rollback, shutdown, escalation, and responsibility should be visible where consequences matter.
-
Test before promotion. Q uses TSLP: Test → Simulate → Learn → Patch, then retest.
-
Think across the lifecycle. Setup, operation, monitoring, updates, recovery, portability, and end-of-life all matter.
-
Count resources. Energy, compute, money, hardware, time, connectivity, and human burden belong in the decision.
-
Preserve the right to leave. Build Your Way In should be paired with Build Your Way Out.
-
State what is not proven. Downloadable, available, impressive, or locally tested does not automatically mean independently validated or production-ready.
Q views and processes
GLEAN2 gathers and compares evidence. LEAP explores scenarios without calling them forecasts. SUPRA adds multi-lens and second-order thinking. Q RX helps prioritize what matters first. T2Y keeps evidence connected to useful decisions. EE adds energy, efficiency, and resource thinking. Human Essence keeps dignity and agency visible. Event Horizon explores emerging risks while keeping action human-gated.
What Q does not publish here
The public site explains the approach but does not expose proprietary prompts, secret thresholds, protected schemas, internal weights, unreleased source code, private security controls, or confidential customer implementations.
How Q builds · Q Builder · Full product index