Skip to content
Furl Docs

Confidence Score

A confidence score measures how reliable a remediation strategy is. It’s a number between 0 and 1 that Furl computes automatically based on the strategy’s track record. Higher confidence means the strategy has a history of successfully fixing the problem.

How the score is computed

Section titled “How the score is computed”

The score combines four factors, weighted by importance:

confidence = (base × 0.2) + (execution_rate × 0.2) + (efficacy_rate × 0.5) + (recency × 0.1)
FactorWeightWhat it measures
Base confidence20%Source quality: curated (0.9), community (0.7), user-created (0.5)
Execution success rate20%How often the strategy runs without errors
Efficacy rate50%How often the fix actually resolves the problem
Recency10%How recently the strategy was used (decays over time)

Why efficacy dominates

Section titled “Why efficacy dominates”

Efficacy gets 50% of the weight because a strategy that runs successfully but doesn’t fix the vulnerability is nearly worthless. A script might exit 0 while the software version doesn’t actually change. Furl detects this through efficacy verification and penalizes the strategy’s confidence accordingly.

How the score is used

Section titled “How the score is used”
  • Governance can set a minimum confidence threshold — strategies below it require human approval.
  • Strategy selection picks the highest-confidence compatible strategy.
  • The score improves automatically as a strategy accumulates successful, verified executions.

This creates a virtuous cycle: the more Furl remediates, the better it gets at picking the right strategy and predicting success.

Section titled “Related”