v3.8.1: trustworthy cloud runtime evidence
Cloud runtime evidence
fallow coverage analyze --cloud now preserves the server's actionability decision, evidence reason, verdict, production classification, freshness, staleness, and source-resolution quality.
When the active deployment has not collected enough isolated production evidence, the CLI keeps the result non-actionable instead of treating any tracked function as sufficient proof. Once the evidence floor is met, the cloud verdict and deployment provenance pass through unchanged. Older cloud responses retain the existing tracked-function fallback.
Full Changelog: v3.8.0...v3.8.1