self_consistency_checker

Cross-artifact consistency validation checks that your spec, theory, implementation, tests, and benchmarks align without contradictions. The self-consistency checker identifies mismatches between declared behavior and actual code structure, data-structure choices and complexity claims, test coverage against requirements, and reported benchmark results against actual outputs, producing a final consistency report.

A consistency checker detects invariant violations where code doesn't match spec, theory-code misalignment in data structures or algorithms, test coverage gaps, benchmark-report incoherence, and broken cross-references. It scans specs, implementations, tests, and reports to surface contradictions that manual review often misses.

Run consistency validation after completing benchmarks and before finalizing reports. It's essential when scaling projects with multiple artifacts—specs, theoretical analyses, implementations, and tests—that must remain mutually coherent. Early detection prevents shipping contradictory claims.

Yes, the checker accesses artifacts via read, grep, and glob operations across specs, theory documents, code files, benchmark data, and reports. It works on any project structure that exposes these artifacts as readable files without requiring specific formats or frameworks.

Prepare all project artifacts—specification documents, theoretical analyses, source code, test files, benchmarks, and reports—in accessible file locations. The checker requires read access to these artifacts to perform cross-validation and generate its consistency report.

Testing validates individual components against their specs; consistency checking validates that specs, theory, code, benchmarks, and reports agree with each other. It catches higher-level contradictions between different artifact types that unit or integration tests won't surface.