bmad-build-auto

Automates one unattended development iteration from intent through planning, implementation, and multi-layer review.

Updated Mar 14, 2026
One-click install
npx skills add https://github.com/ArchaonHW/MingGoRTS --skill bmad-build-auto-archaonhw
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bmad-build-auto
Source: https://github.com/ArchaonHW/MingGoRTS/tree/main/.agents/skills/bmad-build-auto
Command: npx skills add https://github.com/ArchaonHW/MingGoRTS --skill bmad-build-auto-archaonhw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? It removes the need for human supervision during a development cycle by turning a raw intent, story ID, or spec file into a reviewed, verified, and committed code change without asking questions mid-run. ## Core Features & Use Cases - Four-step autonomous workflow: Clarify and route the intent, plan a spec from a template, implement via a subagent handoff, then run layered code reviews with triage and loopback repair. - Multi-layer review pipeline: Runs blind-hunter, edge-case, verification-gap, and intent-alignment reviewers in parallel, then classifies findings into intent_gap, bad_spec, patch, or defer routes. - Spec lifecycle management: Tracks status (draft, ready-for-dev, in-progress, in-review, done, blocked) in YAML frontmatter with change logs, triage logs, and deferred findings. - Use Case: Point it at a story in a stories.yaml epic folder and it produces the spec, implements the change, audits test coverage against the I/O matrix, reviews the diff, and commits the result. ## Quick Start Invoke the bmad-build-auto skill with a description of the change you want implemented and let it run the full unattended plan-implement-review loop.

Frequently Asked Questions about bmad-build-auto

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

FAQPage Schema
How do I run an unattended AI development workflow on a story?

Invoke the skill with a story ID and spec folder, or a freeform intent. It routes the intent, drafts a spec from its template, implements via a subagent, reviews the diff with multiple reviewer layers, and commits the result with status tracking.

How does automated code review triage work in this workflow?

Four review layers (blind hunter, edge-case hunter, verification-gap, intent-alignment) examine the staged diff in parallel. Each finding is verified, graded high/medium/low/false/maybe-false, then routed to intent_gap, bad_spec, patch, or defer with a logged triage entry.

Can I customize the review layers or implementation handoff?

Yes. Override customize.toml values in _bmad/custom/bmad-build-auto.toml or a .user.toml variant. Strings replace defaults, lists append, and review layer tables merge by id, so you can disable layers or substitute external reviewers.

What happens when the autonomous run gets blocked?

The workflow HALTs with an explicit status and blocking condition, writing the result to the spec file or a result file under implementation artifacts. It never asks questions mid-run; unresolved intent gaps, verification failures, or non-converging review loops all produce blocked halts.

Does the workflow require subagents and version control?

Subagents are mandatory; if unavailable the run halts with a blocked status. Version control is expected for baseline diffs and final commits, but the workflow degrades to best-effort change detection and skips committing when it is unavailable.