Establish systematic semantic traceability for invalidated code

Develop and validate a purpose-designed, larger-scale attribution framework that determines whether lines syntactically deleted or replaced after a requirement arrival were semantically invalidated by that requirement, thereby strengthening the interpretation from temporal association toward causal evidence.

Background

The paper's primary invalidation measure identifies deleted or replaced prior agent-authored lines through repository replay, but syntactic disappearance does not demonstrate that the requirement caused a semantic change. An exploratory LLM audit labels only a sample of invalidations and finds that some are incidental. The paper therefore identifies complete semantic attribution as an unresolved measurement issue requiring a dedicated protocol and broader human validation.

References

However, complete semantic traceability remains an open measurement problem that requires a purpose-designed attribution framework to address systematically.

— Requirements After the First Edit: Mining Late Requirement Emergence and Rework in Real-World Coding-Agent Sessions  (2609.03028 - Jiang et al., 2 Sep 2026) in Section 6.3, Threats to Validity; Section 6.5, Future Work