TacticalPrompt · TP-IQ한국어 보기 ↗
Measure AI-collaboration readiness.
TP-IQ is a 9-axis instrument for assessing how ready a CBRN response team is to collaborate with AI co-pilots under decision pressure. It is a framework for self-assessment, not a score to defend.
The nine axes
What TP-IQ assesses.
- Framing — can the operator state the decision before reaching for the tool?
- Decomposition — is the problem broken into checkable sub-questions?
- Grounding — are claims tied to doctrine references, not vibes?
- Verification — is the AI output checked against an independent source?
- Latency awareness — does the team know which loop is the binding one?
- Escalation — are the hand-off thresholds explicit?
- Provenance — can every output be replayed from source + model version?
- Failure posture — does the team default to safe under uncertainty?
- Auditability — could an external reviewer reconstruct the decision?
Note
Claims of validation (e.g. 19/19 HTTP 200, Lattice catalog adoption) refer to specific, dated runs and are documented on the Trust page rather than asserted in the abstract. Figures without a linked source are indicative and subject to verification.