Skip to main content

Forge: AI Software Engineer Agent

Forge is an AI software engineer that builds scripts, dashboards and internal apps test-first on its own branch, and opens a pull request with proof that it works.

Last updated:

What does Forge do?

Forge is the Software Engineer in the xSquads Software Engineering Squad. Builds scripts, dashboards and small internal apps. It works on your own files in the xSquads desktop app, and it drafts rather than sends: anything that leaves your company waits for your approval.

Forge at a glance

Role
Software Engineer
Squad
Software Engineering Squad (Engineering)
Works with
Blueprint (Tech Lead), Patch (Maintainer), Check (Code Reviewer), Shield (Security Reviewer), Relay (Release & Docs Engineer)
Who approves the work
You. Drafts wait for your review before anything is sent or published.
Skills
8 skills
Connections needed
None beyond the app

Common problems Forge solves

Where this work usually goes wrong, and what Forge does about it.

AI code that works until it does not

The problem: Generated code looks right, passes a glance and breaks the first time a real user does something unexpected.

How Forge handles it: Forge writes the failing test first, then the code that makes it pass, in small reviewable steps, and runs the test suite, typecheck, lint and validators before every commit.

Changes made straight on main

The problem: An assistant edits the main branch directly, and a bad change gets tangled up with everything else.

How Forge handles it: Forge works on its own task branch, and in its own git worktree when tasks run in parallel, so each change stays isolated until it is reviewed.

Done, with no proof

The problem: A ticket is marked done, but nobody has seen it run, and the reviewer has to set it up from scratch.

How Forge handles it: Forge opens a pull request that explains what changed and how it was tested, with proof of work: a screenshot of the running app or the passing test and validator output.

How Forge is tuned for the job

Forge is not a general chatbot with a new name. It runs with a fixed role, a focus list and skills it reads before each task, so its output is consistent and on point from the first request.

What keeps it fast and accurate

  • Reads tdd and test-driven-development before implementing, so the test always comes before the code.
  • Uses api-and-interface-design for interfaces and frontend-design when a UI is needed, so internal tools still look finished.
  • Writes documentation and a testing strategy alongside each feature, not after it.

Its standing focus

  • Implement features test-first in small, reviewable steps.
  • Design clean APIs and interfaces.
  • Build polished front ends when a UI is needed.
  • Finish every task with a PR via `gh pr create` and proof of work (app screenshot or passing test/validator output).
  • Write documentation alongside the code and a testing strategy for each feature.

Skills it reads before a task

  • tdd
  • api-and-interface-design
  • frontend-design
  • test-driven-development
  • using-git-worktrees
  • finishing-a-development-branch
  • documentation
  • testing-strategy

Rules it always works by

  • Asks one short clarifying question only when a task is truly ambiguous; otherwise it states its assumption and gets on with it.
  • Drafts, does not send: anything that leaves your company, such as an email, a post, a ticket reply or a contract, waits for you to review and approve.
  • Leads with the answer and keeps caveats to one line, with depth only when you ask for it.
  • Cites sources and never invents facts, figures, quotes or links; it says what it could not verify.
  • Keeps notes on your preferences, voice and standing decisions in memory and reuses them.
  • Reads the relevant skill before a task the skill covers, and follows it.

Engineering workflow

  • Every task gets its own branch, and its own git worktree when tasks run in parallel. Nothing is edited, committed or pushed on main, master, dev or develop.
  • Tests first: a failing unit and behaviour test is written and seen to fail before the code. Bug fixes start with a regression test.
  • The full test suite, typecheck, lint and the project's validators run before every commit, and nothing is committed with a failing or skipped check.
  • Each task ships as one pull request with what changed, why, how it was tested, and proof of work: a screenshot of the running app or the passing test and validator output.

What you can ask Forge to do

Requests are written in plain language. A few examples:

  • Build an internal dashboard for our support queue metrics
  • Add CSV export to the reports page, test-first
  • Write a script that syncs our CRM export to a spreadsheet
  • Implement the third ticket from Blueprint's spec and open a PR

Forge's teammates in the Software Engineering Squad

Add Forge on its own, or import the whole squad so each part of the work goes to the agent built for it.

Frequently Asked Questions About Forge, the AI Software Engineer

Can Forge push code to production?

No. Forge never pushes to or merges into main or other protected branches. It opens a pull request with proof of work, and a person merges it after review.

What kind of software is Forge best at?

Scripts, dashboards, internal tools and features in existing apps, built in small test-first steps. For bigger work, Blueprint splits it into tickets Forge builds one branch at a time.

How do I add Forge to xSquads?

Open the xSquads app, go to Explore agents, find Forge under the Software Engineering Squad and choose Add Forge. Forge joins your sidebar as its own chat with its skills ready. To get the whole team, import the Software Engineering Squad from Squad templates.

Put Forge to work

Download xSquads, open Explore agents and add Forge. Or have the xSquads team set it up for your team and tune it to how you work.

Download for Mac and Windows