architecture-spine
Create, update, or validate a lean architecture spine containing the cross-unit invariants that independently built parts must share. Use before backlog decomposition for multi-epic, cross-system, high-risk, or otherwise coordination-heavy work; not for Task-level code sketches.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- architecture-spine
- license
- MIT
- description
- Create, update, or validate a lean architecture spine containing the cross-unit invariants that independently built parts must share. Use before backlog decomposition for multi-epic, cross-system, high-risk, or otherwise coordination-heavy work; not for Task-level code sketches.
02 资源文件 1
- 参考文档references/spine-template.md1.4 KB
03 SKILL.md 正文
Architecture Spine
Produce a consistency contract, not an exhaustive solution document. Read
../../references/planning-artifacts.md first.
When creating or re-rendering the spine, read
references/spine-template.md.
The inclusion test is strict: if two units one level down built this independently, could they choose incompatibly? Record the decision only when the answer is yes, the choice is non-obvious, and a real trade-off exists. Otherwise defer it or let compliant code own it.
Scope and altitude
Set the scope and altitude from the input:
- initiative spine keeps Features coherent;
- Feature spine keeps Epics coherent;
- Epic spine keeps Stories coherent.
A child spine inherits the parent's decisions as read-only constraints. A conflict is surfaced to the parent; never weakened locally. For brownfield work, inspect the actual repository and adopt existing conventions rather than designing a parallel system.
Coach the load-bearing calls
Default to coaching unless the user asks for speed. Name a coherent design paradigm first, then work through boundaries and dependency direction, state mutation, shared-data ownership, integration contracts, error and security boundaries, deployment/environment invariants, and other dimensions that can create cross-unit divergence. Show realistic alternatives and your lean; the user decides. Verify current versions and fit before naming technologies.
Record decisions and rationale in architecture/.decision-log.md. Stable AD-N identifiers are
never renumbered or reused.
Output
Write architecture/ARCHITECTURE-SPINE.md with:
- scope, altitude, named paradigm, and inherited invariants;
- one block per decision:
AD-N,Binds,Prevents, and enforceableRule; - consistency conventions;
- minimal stack and structural seed, explicitly owned by code after cold start;
CAP-N→ architecture mapping when a product spec exists;- diagrams only where they convey dependency or deployment shape; and
- explicit deferred decisions with the condition that should reopen them.
Keep rationale in the decision log, not the spine. A whole relevant dimension left silent is a gap; a dimension explicitly deferred is a decision.
Review and handoff
Check every AD-N has all three fields, inherited IDs resolve, diagrams are valid, technology
claims are current, capabilities have a home, and the spine contains no decision that could safely
be read from code. Reconcile all load-bearing inputs before finalizing.
On update, append the new decision and re-render while preserving IDs. Validation is read-only and
offers an update. product-spec adopts the spine as a companion. The later architect skill turns
an approved Story into code shapes within this spine; it does not replace or override it.
Adapted from BMad Method's Architecture Spine workflow (MIT, BMad Code, LLC).
04 同仓库更多 Skills
validate-artifact
Validate a single agile artifact (Epic, Feature, User Story, or Task) against the toolkit rule set. Use when the user asks to "validate this story/feature/epic", "check this ticket", "is this artifact ready?", or provides a local draft path or provider item and wants a quality report. Accepts a local draft or a tracker work item. Runs all checks non-blocking and emits a terminal report + persisted report. One artifact per invocation.
split-story
Lateral story-sizing skill. Takes a single User Story and determines if it should be split, how many sub-stories to create, and what split pattern to use — then drafts sub-stories and hands them off. Use when the user asks "split this story", "is this story too big?", "analyze this story for sizing", or provides an artifacts path / tracker id / file path / raw text and wants a split recommendation. One story per invocation.
research-decision
Research a product, market, domain, competitor, user, or technical question in service of a named decision, producing a cited decision brief. Use when planning depends on current evidence rather than more ideation.
product-spec
Distill an idea, brief, PRD, UX contract, architecture spine, issue, or mixed notes into a compact canonical product contract with stable capabilities and companions. Use before backlog decomposition when intent is defined; update or validate an existing product spec through the same skill.
product-requirements
Create, update, or validate a right-sized Product Requirements Document with stable requirements, explicit scope, user journeys where useful, and measurable outcomes. Use for multi-person or multi-epic alignment; skip it when a concise product spec is enough.
product-brief
Create, update, or validate a concise product brief that captures a clear concept, its users, value, scope, and success criteria. Use when the user wants a one- or two-page product narrative before a PRD or product spec.