A deterministic rule-based monitor that detects coding-agent drift, paired with an LLM advisor called only when needed, raises SWE-agent issue resolution rates by up to 15.2 percentage points at about $0.08 extra per task.
Introducing claude opus 4.7,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.SE 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Online Monitoring and Corrective Steering of Programming Agents
A deterministic rule-based monitor that detects coding-agent drift, paired with an LLM advisor called only when needed, raises SWE-agent issue resolution rates by up to 15.2 percentage points at about $0.08 extra per task.