Fetching from the wire…
01
02
03
04
05
06
07
08
Research2026-08-13 · source-backed
arXiv 2608.11386 ran 11,700 repository-issue-fixing trajectories across six tool architectures with capabilities held roughly equal. Structured interfaces improved run-to-run consistency up to 4.7x, natural-language search tools raised relevant-file discovery over 11%, and Python CodeAct-style interfaces cut steps 41.6% and tokens 56.3%. Cognitive-scaffolding tools like reasoning logs showed almost no effect. The highest-leverage knob on your coding agent is the tool schema, not the system prompt.
Each link below shares sources, entities, or timing with this story.
Shared entity: Tool / Same source domain / Shared topic / Earlier coverage / Tension
Both cover Tool; reported by the same outlet (arxiv.org); overlapping topics (agent, coding, tool).
Shared entity: Tool / Shared topic / Earlier coverage / Tension
Both cover Tool; overlapping topics (agent, prompt, tool); earlier Tool coverage from 2026-02-21.
Shared entity: Tool / Same source domain / Shared topic / Earlier coverage
Both cover Tool; reported by the same outlet (arxiv.org); overlapping topics (agent, beat).
Both cover Tool; reported by the same outlet (arxiv.org); overlapping topics (agent, discovery).
Both cover Tool; reported by the same outlet (arxiv.org); overlapping topics (agent, tool).
Same source domain / Shared topic / Tension
Reported by the same outlet (arxiv.org); overlapping topics (agent, beat, capability, tool); pushes against this story (against).
Shared entity: Structured / Shared topic / Earlier coverage / Tension
Both cover Structured; overlapping topics (agent, tool); earlier Structured coverage from 2026-03-25.
Both cover Structured; overlapping topics (coding, engineering); earlier Structured coverage from 2026-03-14.