bmad-spec

Distill intent inputs into a five-field SPEC.md kernel with companion files and append-only memlog.

Updated Jul 2, 2026
One-click install
npx skills add https://github.com/tuanpa-nhg-eng/nhg-ipms --skill bmad-spec-tuanpa-nhg-eng
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bmad-spec
Source: https://github.com/tuanpa-nhg-eng/nhg-ipms/tree/main/.claude/skills/bmad-spec
Command: npx skills add https://github.com/tuanpa-nhg-eng/nhg-ipms --skill bmad-spec-tuanpa-nhg-eng

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) components.

What problem does it solve? Turning vague ideas, PRDs, transcripts, or mixed multi-source inputs into a single canonical, preservation-validated specification that downstream planning and development skills can consume without losing load-bearing decisions. ## Core Features & Use Cases - Spec Kernel Distillation: Converts any intent input into SPEC.md carrying the five-field kernel (Why, Capabilities, Constraints, Non-goals, Success signal) with stable capability IDs. - Companion & Memlog Management: Routes overflow content into named companion files and records every decision in an append-only .memlog.md from which the spec is re-derived, preventing merge drift. - Validation & Story Breakdown: Runs two-pass self-validation (coherence and preservation) against eight Spec Law rules, and optionally breaks the spec into a stories.yaml dispatch file. - Use Case: Paste a messy customer email thread and ask for a spec; the skill produces a spec folder with SPEC.md, a glossary companion, and a memlog recording every decision and open question. ## Quick Start Ask the assistant to distill the attached PRD or your pasted idea into a BMad spec with capabilities, constraints, and open questions.

Frequently Asked Questions about bmad-spec

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

FAQPage Schema
How do I turn a PRD or brain dump into a structured spec?

Provide the file path or paste the content and ask for a spec. The skill distills it into SPEC.md with a five-field kernel (Why, Capabilities, Constraints, Non-goals, Success signal), routing overflow detail into named companion files.

What is the difference between SPEC.md and the memlog?

The .memlog.md is the canonical append-only record of every decision, capability, and assumption. SPEC.md is derived from the memlog on each run and is never hand-edited, so updates re-render the spec without merge drift.

Can multiple skills update the same spec over time?

Yes. Invoking with the same slug lands on the existing spec folder and updates it in place. Capability IDs are preserved, new capabilities get the next unused CAP-N, and retired IDs are never reused.

When should content go into a companion file instead of SPEC.md?

Content needing more than one kernel-shaped line goes into a companion: multi-item catalogs, tables, diagrams, glossaries, and long-form references. Diagrams always land in a companion since the kernel holds prose only.

What happens if my input is too vague to distill?

If the input is genuinely too thin, the skill stops and suggests running bmad-prd first to draw out the vision. In headless mode it returns a blocked JSON response with error_code insufficient_intent instead of inventing content.

How does story breakdown work in bmad-spec?

In interactive mode with SPEC.md on disk, the skill walks capabilities with you and proposes one story per independently shippable slice, writing stories.yaml. Headless runs never perform story breakdown.