Skip to content
IFAI

AI tools

Google brings Jules coding agent to your terminal and CI/CD pipelines

Published Reach 48

Google just launched Jules Tools, a lightweight CLI bringing its AI coding agent straight into your terminal, alongside early access to the Jules API. Developers can now wire Jules into Slack bugs or CI/CD pipelines. Google also rolled out persistent memory for coding preferences, a file selector, and environment variable controls.

Google brings Jules coding agent to your terminal and CI/CD pipelines
Image: blog.google

Google is bringing its autonomous coding agent straight into your terminal.

On Thursday, Google introduced Jules Tools, a new lightweight command-line interface, and officially opened public access to the Jules API. The releases allow developers to run, manage, and verify coding tasks right inside their local terminals, CI/CD pipelines, and chat apps like Slack, without having to switch tabs to GitHub or a web browser.

Less chatting, more delegating

Jules, which runs on Google's Gemini 2.5 Pro model under the hood, was previously only accessible through a web interface and GitHub. With Jules Tools, developers can trigger tasks alongside regular bash commands and stay focused in their local development environment.

Google already offers Gemini CLI, but the team says the two serve entirely different purposes. Kathy Korevec, director of product at Google Labs, told TechCrunch that Jules Tools handles tightly scoped tasks independently. Google senior developer advocate Denise Kwan added that while Gemini CLI requires continuous back-and-forth iteration, Jules plans the job, gets user sign-off, and executes in the background on its own, only interrupting if it gets stuck.

The newly public API was previously used internally by Google. Developers can now wire Jules directly into their own systems, such as triggering an agent run whenever a bug is reported in Slack or baking it into automated build pipelines. Over recent weeks, Jules also gained memory for user preferences, file selectors for tighter context, and environment variable management.

Jules offers a free tier with up to 15 tasks per day and 3 concurrent tasks. Power users can bump limits via Google AI Pro ($19.99/month) or Ultra ($124.99/month) subscriptions for roughly 5x and 20x the capacity.

Why it matters

For working software engineers, context switching kills productivity. By shipping a CLI and an API instead of forcing people into a standalone browser interface, Google is letting Jules slot directly into existing terminal muscle memory.

We think Google is making the right strategic bet by leaning into asynchronous delegation rather than just autocomplete. Unlike vibe coding toys designed for non-technical users, Jules acts like a junior developer tackling PRs, fixing broken tests, and handling tedious fixes. In fact, Korevec noted that users are already bringing projects to Jules when simpler vibe coding tools hit their ceiling.

The friction point right now is monitoring. Asynchronous agents need reliable nudges when they get stuck, but Jules currently lacks native mobile notifications. If a pipeline halts while an engineer is away from their desk, the task just sits there waiting. Still, meeting engineers inside the terminal is the fastest way to win real adoption.

Sources (2)

  1. [1]
    blog.google(opens in a new tab)Originalblog.google
  2. [2]
    techcrunch.com(opens in a new tab)Reporttechcrunch.com

Best of AI, daily