brainstorming

Guides collaborative design dialogue that classifies requests and produces approved specs before implementation.

Updated Aug 28, 2026
One-click install
npx skills add https://github.com/jeremybrasher/grokbot-skills --skill brainstorming-jeremybrasher
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: brainstorming
Source: https://github.com/jeremybrasher/grokbot-skills/tree/main/skills/brainstorming
Command: npx skills add https://github.com/jeremybrasher/grokbot-skills --skill brainstorming-jeremybrasher

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? It prevents wasted implementation work by forcing intent exploration, requirement clarification, and explicit human approval before any code is written for new features or behavior changes. ## Core Features & Use Cases - Three-Path Classification: Routes every request into spike, bounded, or architectural paths with a one-way upgrade ratchet when hidden complexity appears. - Approval Gates and HOLD Table: Blocks implementation until the human partner approves the design, and returns explicit gap statements instead of guessing when preconditions are missing. - Spec and Plan Authoring: Produces reviewed design specs and executable implementation plans with file maps, per-task test cycles, and no placeholders. - Optional Visual Companion: Serves browser-based mockups and diagrams via a local WebSocket server for questions that are genuinely visual. - Use Case: Before building a new API endpoint, classify the request as bounded, read the existing flow, present a short in-chat design, and implement only after the partner says yes. ## Quick Start Ask the agent to brainstorm the design for a new feature before writing any code, and answer its clarifying questions one at a time.

Frequently Asked Questions about brainstorming

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

FAQPage Schema
How do I brainstorm a feature design before writing code?▼

Invoke the skill before any creative work and it will classify the request as spike, bounded, or architectural. It then asks clarifying questions one at a time, presents a design, and waits for your explicit approval before any implementation begins.

What is the difference between spike, bounded, and architectural paths?▼

A spike answers a feasibility question with throwaway investigation, bounded covers a well-scoped change to an existing flow in the repo, and architectural covers new projects or subsystem restructuring. Each path scales its artifacts but never skips the approval gate.

When should I use the visual companion browser tool?▼

Use it only when a question is genuinely visual, such as comparing layouts, mockups, or diagrams. Conceptual or requirements questions stay in the terminal, and the companion is offered just-in-time as its own message, never upfront.

Can the skill start coding right after presenting a design?▼

No. A hard gate blocks all implementation, scaffolding, and implementation-skill invocation until the human partner explicitly approves the presented design. Presenting a design and starting work in the same turn is treated as skipping the gate.

What happens if requirements are missing or the repo is unreadable?▼

The skill returns a HOLD instead of a design, stating exactly what is missing and what is needed. It refuses to invent a product, scope, or classification when the object of work, an existing flow, or an approving partner is absent.