Conflation / provenance-honesty scorecard

Name: Netra Fused Agent Kernel
Author: Anthony Chaudhary

Auto-generated by tools/conflation_scorecard.py. Do not hand-edit; re-run the tool.

Grade A - score 100.0 - conflation_debt 0 - 2 reporting surface(s) - 6 external-value string(s)

The law: every reported number/status labels its provenance - WITNESSED (a fact fak authored) vs OBSERVED (a value fak relays from an external party) - and a bad OBSERVED value is never blamed on a fak action unless a witnessed signal proves the fault is fak’s.

KPI	score	detail
provenance_labeled	100	6/6 external-value strings carry a provenance label
no_false_attribution	100	no observed-miss-blamed-on-fak prose
fault_signal_isolated	100	fault signal isolated where families mix provenance

fak conflation scorecard - the provenance-honesty stick

fak's deterministic conflation scorecard: does every number and status fak reports label its provenance - WITNESSED (a fact fak authored) vs OBSERVED (a value relayed from an external party) - folded into a conflation-debt count.

Conflation / provenance-honesty scorecard