Terminal Coding Agents

A coding agent takes an instruction in plain language, then reads files, edits them, runs commands and checks the result in a loop until the task is finished. Agents are built on large language models and became practical once those models could call tools reliably.

 

In the terminal an agent works inside a repository: it applies a patch, runs the test suite, reads the failure and tries again, with Git providing the review and the rollback for whatever it changed. Most ask permission before running a command, read project instructions from a file checked into the repository, and connect to other services over the Model Context Protocol.

 

Below is a list of coding agents that run in the terminal.