Run an unattended agent factory that turns GitHub issues into pull requests

Abant ran agents with nobody in the session for four months: they took GitHub issues, worked in isolated worktrees and opened PRs reviewed by another agent; about 14 percent had a failed CI run.

Done withCustom agentinGitHub· Nov 15, 2026

What they did

Eduardo Rosado describes the software factory his company Abant has run on its own product from July to November 2026. Agents pick up GitHub issues, work in isolated git worktrees with no person watching, and open pull requests that a different agent reviews. Each issue lists the files the change may touch and acceptance criteria that end with the repo's validation command, and a plan cannot be published as issues without both. An orchestrator compares the files a PR changed against the issue's list and holds the merge for a person if anything falls outside it. A session that finds the spec ambiguous prints one concrete question and exits; the task waits until someone answers. Per-file size and coupling budgets fail validation when exceeded. The agent products and models are not named.

How it went

Every session is told to run full validation before opening a PR, yet roughly one in seven PRs (about 14 percent) had at least one failed CI run. Rosado treats that as an upper bound, since some failures were flaky tests or differences between the agent's machine and CI, and the causes were not classified. His conclusion is that instructions alone did not keep changes green and the checks caught what slipped through. Nothing in the factory catches a change that is bigger than it needed to be.

Worth knowing

The useful split is between rules a machine can check (allowed files, acceptance commands), rules you can only make cheap to follow (stopping to ask), and rules left to the reviewer (simplicity). File-level scope checks still let an unneeded edit through inside an allowed file. The post is in Spanish, and the mix of models over four months means it shows no per-model trend.

Try it

With Custom agent· 4 parts to fill in

Take GitHub issue [issue number] in [repo]. Work in a fresh worktree. Change only the files the issue lists as allowed; if the job needs any other file, stop and say which and why. If the spec is ambiguous, print one specific question and stop without guessing. When done, run [validation command] and fix failures until it passes, then open a pull request that names the acceptance criteria and how each was checked. Do not merge; a separate reviewer and [person or rule] decide that.

Square brackets are the parts to fill in. The prompt is ours; the story is theirs.

Where they said it

Related