Jules
Jules is Google's asynchronous AI coding agent. Given a task description, Jules opens a GitHub repository, makes changes across multiple files, runs the project's test suite, and returns a pull request — without requiring the developer to maintain an active session. The agent works in the background; the developer reviews the PR when it is ready.
Most AI coding agents — Claude Code, GitHub Copilot CLI, OpenAI Codex CLI — operate in an interactive, synchronous model: the developer initiates a task and monitors the session. Jules adopts an asynchronous model: submit a task, do other work, review the result. This is closer to delegating to a junior developer than to pair programming with an AI.
Placed in Assess at inner position because the async execution model is a meaningful differentiation in a crowded coding agent space. Teams with large backlogs of well-scoped tasks — bug fixes, test additions, refactors — may find async agents more practical than interactive ones. The inner position reflects that the model is worth actively evaluating now rather than waiting for further maturation. Jules is Google's entry in this space; similar async patterns are emerging in other agents.
The key evaluation criterion is task definition quality: async agents succeed when the task is specific enough that the agent can complete it without mid-task clarification. Vague tasks produce poor PRs regardless of model quality.
