Assess threat on an evidence-tiered rubric

The challenge

Threat assessments mix a single blurred frame with weeks of confirmed observation and present both as findings.

The scenario

Several signals suggest increased pressure in one sector, and the team must decide where to deploy limited patrols.

What goes in

Field signals

Observations, reports and imagery from the area.

Historical context

Prior incidents and patterns.

What you can ask

Illustrative prompts. Results depend on your own data and are not deterministic.

• “Where has threat increased this month?”Copy
• “What evidence supports this score?”Copy
• “Which parts of this are inference rather than observation?”Copy

How Clearception approaches it

How the work is structured. Each step shows what its output is grounded in.

1. 1

   Gather signals

   Source

   Collect field observations and reports with provenance.

2. 2

   Tier the evidence

   Observed

   Classify each signal by how strong its basis is.

3. 3

   Score by component

   AI-inferred

   Apply the rubric so weak evidence caps the score.

4. 4

   Separate status

   Human-verified

   Mark observation, inference and report distinctly.

What you get

The shape of what comes back — not a promised result.

• Tiered threat assessment

  AI-inferred

  A score decomposed by component, with evidence strength per element.

Why it matters

Patrols follow evidence

Limited capacity goes where the basis is strongest.

No false certainty

A thin signal cannot produce a confident score.