kiro-verify-completion

Verifying a bugfix with fresh evidence requires running the canonical validation command against current code state and checking the fresh command output and exit codes. The result is VERIFIED, NOT_VERIFIED, or MANUAL_VERIFY_REQUIRED with any coverage or integration gaps reported.

Preventing false success reports requires enforcing a deterministic verification gate that identifies the exact completion claim, runs canonical validation commands, and checks evidence-to-scope alignment before accepting any task, fix, test_or_build, or feature_go claim as complete.

Scope-aligned evidence means the fresh command output and exit codes directly map to the feature_go claim, including integration checks, coverage checks, and runtime smoke checks that confirm the feature is ready for a GO decision without relying on stale or secondhand reports.

Completion verification escalates to MANUAL_VERIFY_REQUIRED when mandatory checks cannot be performed, such as when canonical validation commands are unavailable or cannot produce the fresh output and exit codes needed to confirm task, fix, test_or_build, or feature_go claims.

Yes, completion verification applies tailored checks for TEST_OR_BUILD claims during CI, requiring fresh command output and exit codes from canonical validation commands to produce VERIFIED or NOT_VERIFIED results with evidence and gaps.

The best way to enforce quality assurance gates is to apply claim-specific rules for TASK, FIX, TEST_OR_BUILD, and FEATURE_GO scenarios, requiring canonical validation commands, fresh output, exit codes, and evidence-to-scope alignment checks before producing a final verification outcome.