validate-completion

Software task verification works by either executing the feature to confirm expected database state changes and graceful failures, or inspecting artifacts for non-empty files, correct imports, and tests exercising real code paths. This confirms completion claims are backed by actual results.

To validate a feature is complete, invoke the feature to confirm expected database state changes and graceful failure handling. Alternatively, inspect artifacts rigorously for non-empty files, absence of placeholders, correct imports, and tests exercising real code paths.

Execution mode invokes the feature for real to check for graceful failure and confirm expected database state changes. Inspection mode rigorously inspects artifacts for non-empty files, no placeholders, correct imports, and tests exercising real code paths.

No, execution capabilities are not strictly required. The verification process requires either execution or inspection capabilities. If execution is not possible, you can rigorously inspect artifacts for non-empty files, correct imports, and tests exercising real code paths.

When inspecting software artifacts for task verification, check for non-empty files, absence of placeholders, and correct imports. Additionally, ensure that tests exercise real code paths to confirm the task or feature is fully functional.

Auditing task completion claims prevents unverified features from being marked as done. By executing the feature or inspecting artifacts rigorously, you ensure the goals of a task are genuinely achieved and backed by actual results, maintaining software quality.