conflation-score

Provenance-honesty in fact-reporting ensures metrics accurately distinguish between witnessed internal signals and observed external data. It prevents truth-maintenance bugs where a system incorrectly attributes external provider-side errors to its own internal actions, eroding trust in reported metrics.

To audit reporting surfaces for provenance labeling debt, run the internal Go-based conflation-scorecard verb. This scorecard measures debt within the fak kernel and verifies the integrity of the reporting boundary to identify where external values require proper labeling.

To fix misattributed provider-side issues in your agent kernel summaries, apply attribution correction to identify prose that misattributes upstream failures. Enforce a strict distinction by correctly labeling external values as OBSERVED to prevent false blame on the kernel.

Yes, provenance auditing requires the internal Go-based conflation-scorecard verb to operate. It functions specifically within reporting surfaces inside the fak kernel to measure debt and verify the integrity of the reporting boundary.

You should run a conflation scorecard audit after adding new metrics or exit summaries to your reporting surfaces. This ensures external values are correctly labeled as OBSERVED, preventing false blame on the kernel for upstream failures.