Executable evaluation remains the baseline for this period. The strongest claims come from model-external pieces: SWE-Edit’s read/write split, Agentic Harness Engineering’s rollout-driven harness edits, and SAFEdit’s…
The concrete openings are in the code-editing path around the model: narrower read results, dedicated patch execution, test-backed repair loops, versioned harness changes, and ranked reports from uncovered code.
The strongest signal for this day is that coding research is tightening around evidence that can be checked in real repositories.
Current coding-agent work points to three concrete changes: add executable specification checks to pull-request review, add repository-context reasoning traces to evaluation, and route cross-file follow-up edits through…