Skip to main content

Crate ucm_reason

Crate ucm_reason 

Source
Expand description

Reasoning & Test Intent Engine — the “brain” of autonomous QA.

This crate answers the core questions:

  1. What changed? (Change analysis)
  2. What is impacted? (Impact analysis with graph traversal)
  3. What is NOT impacted? (Equally important — explains WHY)
  4. What should be tested? (Test intent with confidence tiers)
  5. What is uncertain? (Ambiguity detection)
  6. WHY did the system reach these conclusions? (Explanation chains)

Every output includes an explanation_chain — a sequence of ReasoningStep structs, each with evidence (what data), inference (what conclusion), and confidence (how sure). This is the “show your work” layer that makes the system auditable.

Re-exports§

pub use ambiguity::*;
pub use explanation::*;
pub use impact::*;
pub use intent::*;

Modules§

ambiguity
Ambiguity detection — flags conflicts, drift, and missing data.
explanation
Explanation engine — traceable reasoning chains for every conclusion.
impact
Impact analysis — determines what is directly/indirectly impacted by a change.
intent
Test intent generator — produces test recommendations from impact analysis.