auto-manager

Orchestrate epics by filing dependency-ordered sub-issues and driving parallel auto-engineer subagents.

5|1|Updated Apr 14, 2026
One-click install
npx skills add https://github.com/dburkart/auto-engineer --skill auto-manager
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: auto-manager
Source: https://github.com/dburkart/auto-engineer/tree/main/templates/skills/auto-manager
Command: npx skills add https://github.com/dburkart/auto-engineer --skill auto-manager

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

auto-manager turns a big, fuzzy piece of work into an execution-ready epic by discovering what exists in the backlog and codebase, filing a dependency-ordered plan, and then driving multiple auto-engineer subagents until everything merges.

Core Features & Use Cases

  • Epic scoping from fuzzy topics or an epic issue: accepts either a concrete parent issue or a topic string, then generates a user-approved scope before making tracker changes.
  • Tracker-agnostic sub-issue filing and refinement: creates well-scoped sub-issues via the sibling /file-issue skill and iterates with the user (rescope, fold, drop, add missing) using the tracker playbook recipes.
  • Parallel workstream planning and orchestration: partitions work into independent workstreams, then spawns /auto-engineer subagents in isolated worktrees to ship and merge the epic autonomously.

Quick Start

Use the auto-manager skill to manage an epic by running: /auto-manager #18

Frequently Asked Questions about auto-manager

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

FAQPage Schema
How do I orchestrate a GitHub epic end-to-end with parallel subagents?

Orchestrating a GitHub epic end-to-end with parallel subagents involves discovering backlog and codebase context, filing dependency-ordered sub-issues, then driving isolated auto-engineer subagents to merge results through CI gated workflows.

What is the best way to convert a fuzzy topic into a dependency-ordered backlog plan?

Converting a fuzzy topic into a dependency-ordered backlog plan requires generating a user-approved scope before making any tracker changes, then creating well-scoped sub-issues that can be iteratively refined or dropped using tracker playbook recipes.

Can I use auto-engineer subagents in isolated worktrees for parallel workstream planning?

Yes, you can use auto-engineer subagents in isolated worktrees for parallel workstream planning. The system partitions work into independent workstreams and spawns subagents with completion-based respawning to ship and merge the epic autonomously.

Does autonomous software delivery require CI and review gates for merging sub-issues?

Autonomous software delivery requires CI and required review gates for merging sub-issues. The orchestration process enforces these gates while applying phased guardrails to prevent destructive tracker writes during the initial discovery phase.

What are the limitations of backlog triage during epic scoping?

A limitation of backlog triage during epic scoping is that strict phased guardrails prevent destructive tracker writes during discovery. You must generate a user-approved scope before any issues are filed or tracker changes are made.