Agentic development uses AI agents to plan and execute multistep programming tasks, moving beyond autocomplete toward collaboration with limited human intervention. section 8.6
Agents can locate bugs, update deprecated APIs, generate patches and tests, refactor functions, and pursue high-level goals across entire codebases. They may also reason diagnostically, use command-line tools, remember prior sessions, and collaborate as multiple agents. section 8.6
Two broad forms illustrate the shift:
- Interactive IDE agents work alongside developers in real time, editing code, testing changes, and maintaining context. section 8.6
- Autonomous background tools operate more independently for extended periods, raising practical questions about oversight and reliability. section 8.6
Delegating a feature
A developer can ask an agent to “add OAuth support.” The agent may chain edits across files, generate tests, run them, and iterate with minimal prompting. section 8.6
Figure 8.3 Workflow of an AI coding agent embedded in an IDE, from task planning to deployment
Follow the workflow from task planning through deployment to see how an IDE-embedded coding agent organizes delegated work. section 8.6
