The day’s strongest software-agent papers make large language models (LLMs) propose code, plans, or actions, then check them with execution, verifiers, retrieval gates, or live tools.
Three practical changes stand out: enforce authorization inside retrieval and tool loops for shared enterprise agents, route new-service creation through approved Backstage templates, and compile repeated…
The day’s strongest theme is control around LLM coding systems. New work tries to cut prompts to the code that matters, trace what crosses the natural-language/program boundary, and fence agent actions with tighter…
Concrete work is forming around three control points in LLM coding systems: shrinking repository context before repair, restoring taint and slicing across LLM call boundaries, and limiting what coding agents can touch…