The agent coding Jules Google goes beyond chat with a new command line and the API interface

Google He wants his coding assistant to be much more integrated with programmers than ever. The company desires to make it a more native tool, hoping that more people will use it outside the chat interface.

Jules, which the company announced for the first time in December 2024It will gain two new functions: Jules’s API to facilitate the integration of IDE and CLI Jules tools, enabling the agent to open directly on the command line.

More corporations consider that bringing their agents, focused on coding or not, to the application people remove many friction for corporate users. Jules moves this trend a step further, adopting the same flow as programmers.

- Advertisement -

“To this day you have entered Jules in your web browser, but we know that programmers live in the terminal,” said Kathy Korevec, product director at Google Labs, In Blog post. “This is where we test, build, debug and send. That is why we built Jules Tools, a light command line interface, so you can shoot tasks, check what Jules does, and make an agent, all without leaving work flow.”

Through Jules Cli and API, Google said that enterprises would receive “greater control and flexibility, where and how you can use Jules.”

Google in May He released Jules on beta and announced the integration of Github. Happened generally available In August with higher limits rates for Google AI Pro and Ultra users.

Departure

As Korevec said, the fundamental way in which programmers and enterprises interact with agents is the chat interface.

Other coding agents began to integrate their Coding tools and CLIs. Openai implemented the refined version GPT-5, called GPT-5-Codex, which can initiate the strategy of unifying its Code coding assistant with IDES, CLID and other work flows. Google also Gemini cli was releasedwhich works similarly to the CLI of Jules’ tools, but is open and may be transferred to other platforms.

Coding agents are already becoming obligatory tools for enterprises, and as them Preferences solidifysuppliers like Google, OpenAI i Anthropic They want these agents to be more mind for users.

Enterprises imagine AI agents more passive, the tools of the environment, Which may be difficult to assume whether most individuals enter them through chat and must make them work. If the adoption is high, that a more proactive and integrated future will likely be clearer.

How do the API tools and interface work

Developers can install Julesa tools via NPM, which can then print a guide on how one can use it. During your stay at the CLI, the engineer or coder can use the code command to mount Jules to perform the task, and the code flag will adjust it. For example, this string, Jules -Light, will go to light mode.

On the API side, enterprises can connect the API of Julesa with other used platforms. For example, they will connect it to Slack so that some team members can free the tasks directly from Slack if an error is reported, which can then affect their CI/CD pipeline.

Google also added other updates to cut back the delay and fix some problems with the environment and the file system. These include:

  • File selector to call specific chat files so as to add context

  • Memory that offers Jules the opportunity to recollect preferences in the future

  • Management of environmental variables, which Jules gives access to those variables when performing the task

Answer so far

Since the announcement, the answer was mostly positive. However, some are confused with two offers of the Google CLI CLI agent.

Latest Posts

Advertisement

More from this stream

Recomended