automation-builder

name	automation-builder
description	Turns an approved build spec into a minimal technical implementation plan, including system components, human-in-the-loop boundaries, workflows, and delivery milestones. Use when an internal automation, concierge workflow, or lightweight product has already been chosen and now needs a concrete build plan.
compatibility	Works best when the input already includes a selected wedge or build spec.
metadata	{"author":"opencode","version":"0.1"}

Automation Builder

Overview

This skill converts a chosen workflow or product wedge into a minimal technical implementation plan. The goal is to build the smallest reliable system that fits the chosen wedge, not a future-proof platform.

Load references/system-patterns.md when choosing architecture shape. Load references/human-loop-rules.md when deciding what stays manual.

When to use

• When a build spec already exists
• When a concierge workflow or internal tool needs technical planning
• When a validated wedge is ready for a lightweight app or automation

Do NOT use when:

• The business wedge is still unclear
• The task is still opportunity selection rather than implementation planning

Workflow

1. Define the system objective

Name:

• primary user
• primary trigger
• primary output
• human escalation point

2. Choose minimal architecture

Prefer the simplest architecture that can ship.

Define:

• components
• data inputs
• outputs
• storage
• integrations

3. Keep human-in-the-loop where needed

Be explicit about:

• what stays manual
• what is reviewed by a human
• what is safe to automate now

4. Produce the build plan

Return both:

• a readable build plan
• a JSON block using assets/build-plan-template.json

Use the template field names exactly. Populate system_objective.primary_user, system_objective.trigger, system_objective.output, and system_objective.human_escalation explicitly.

Default sections:

• System Objective
• Architecture
• Human Loop
• Milestones
• Out Of Scope
• Structured Build Plan

Checklist

• The plan is minimal, not platform-shaped
• Human escalation is explicit where needed
• Inputs, outputs, and components are concrete
• Milestones are sequenceable
• Readable plan and structured JSON are both present
• At least one manual step, one automated step, and one escalation reason are explicit

Key principles

1. Minimal architecture first — Build for the wedge, not the imagined company.
2. Safety before full automation — Human review belongs anywhere the workflow is high-risk or still ambiguous.
3. Out-of-scope matters — What not to build is part of the plan.

Optimization Notes

• Preserve the user's requested output shape exactly and do not substitute generic advice for concrete artifacts.
• Include exact commands, code structures, protocol fields, tags, parameters, file paths, or deliverable sections when the task asks for them.
• Make safety gates explicit before irreversible, destructive, externally visible, or compliance-sensitive actions.
• For multi-step work, present steps in execution order and include validation or rollback checks where relevant.
• Avoid overfitting to a single eval example: express lessons as reusable rules, not as task-specific answers.