agentic coding
Agentic coding is a software development approach in which an AI agent writes, runs and revises code across multiple steps on its own.
Agentic coding is a software development approach in which an AI agent takes a developer's goal, writes code, checks the results and errors of running it, and continues making revisions across multiple steps on its own. Unlike code autocompletion, where a person accepts suggestions line by line, the agent directly performs tasks such as editing files, running commands and executing tests.
The developer defines the scope and acceptance criteria of the work and reviews what the agent produces. The concept applies to coding agents that combine large language models with tool-calling capabilities.
This entry is based on AIPOST articles and widely known facts. If something is wrong, please send us a correction request.
Articles covering this entry
Claude Sonnet 5.5 costs half as much as Claude Opus 5.5, yet it scored 70.6% on Terminal-Bench 4.0, an agentic coding benchmark, ahead of Opus 5.5 at 64.4%.
…dels work as everyday workhorses, and Sonnet 5.5 is especially strong at agentic coding, where an AI agent writes and revises code on its own, and at orchestrat…
Claude Opus 5.5 is built for agentic coding, in which a model uses tools to carry out multistep development tasks.