auto-engineer

Automate GitHub issue delivery from planning through merge with CI and review handling.

5|1|Updated Apr 14, 2026

Search Tags:#autonomous development #issue triage #ci polling #github pull requests #sdlc automation #closed-loop delivery #review resolution

Ready to run

npx skills add https://github.com/dburkart/auto-engineer --skill auto-engineer

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: auto-engineer
Source: https://github.com/dburkart/auto-engineer/tree/main/.claude/skills/auto-engineer
Command: npx skills add https://github.com/dburkart/auto-engineer --skill auto-engineer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires gh, git, tmux, mcp__github__create_pull_request, mcp__github__add_issue_comment, mcp__github__merge_pull_request.

What problem does it solve?

It eliminates the manual, multi-step effort required to take an unblocked issue from planning through implementation, CI validation, review resolution, and merging.

Core Features & Use Cases

End-to-end autonomous SDLC loop: pick → plan → implement → open PR → poll CI/review → merge, then repeat across cycles.
Policy-driven issue selection: filters blocked or unready issues and prioritizes by label and project playbooks.
CI and review handling with guardrails: polls for terminal CI states, stops on action-required findings, and applies limited auto-fixes for format/lint only while filing follow-ups when needed.

Quick Start

Use auto-engineer by asking your assistant to run the loop for the next unblocked issue, for example: run /auto-engineer.

Frequently Asked Questions about auto-engineer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I automate a closed-loop GitHub pull request pipeline from issue triage to merge?▼

The autonomous SDLC loop operates by polling CI for terminal states, applying limited auto-fixes for format and lint issues, and filing follow-up issues when needed. It enforces safety stops on action-required checks or unsafe conditions to prevent uncontrolled merges.

Do I need the GitHub CLI to run autonomous software delivery workflows?▼

You can trigger the loop by asking your assistant to run it for the next unblocked issue, such as running /auto-engineer. The skill reads SDLC playbooks for conventions, filters blocked or unready issues, and prioritizes by label to begin the implementation cycle.

What happens when CI checks require manual action during an automated PR pipeline?▼

For format and lint failures, the loop applies limited auto-fixes and files follow-up issues when needed. It continuously polls for terminal CI states to determine if the pull request is ready to merge or if further review resolution is required.

Can I use autonomous development to triage and resolve GitHub issues without manual intervention?▼

The workflow handles review resolution by addressing feedback automatically. It merges pull requests when CI is green, then repeats the cycle across multiple issues to maintain continuous iteration over your GitHub-based engineering workflows.