Skip to content
Agentmart
Submit a tool

ag-brainstorming

Use when starting creative work, building components, or modifying behavior.

Skill by ngxccc

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/ngxccc-ag-brainstorming

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use when starting creative work, building components, or modifying behavior. Trigger keywords: brainstorm, spec, design, approach, requirements, clarify. Explores user intent, requirements, architecture trade-offs, and design specs before implementation.

Use cases: design

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/ag-brainstorming

What's inside

Skill name
ag-brainstorming
Description
Use when starting creative work, building components, or modifying behavior. Trigger keywords: brainstorm, spec, design, approach, requirements, clarify. Explores user intent, requirements, architecture trade-offs, and design specs before implementation.
License
MIT
Instructions
73 lines in SKILL.md · view on GitHub

Files (3)

  • SKILL.md
  • brainstorming-guide.md · references/
  • design-spec-template.md · references/

SKILL.md (first lines)

# Brainstorming Ideas Into Designs (`ag-brainstorming`)

## When to Apply

This skill is designed for the agent to reference when:

- Starting any creative work, building new components, adding features, or modifying system behavior.
- Exploring user intent, requirements, architecture trade-offs (Trade-off Matrix), and design specifications before implementation.
- Evaluating designs against the 7 Baseline Engineering Evaluation Domains (Domain A–G: Security, UI/UX, Performance, Reliability, Maintainability, Observability, Compliance).
- Triggered by explicit keywords: `brainstorm`, `spec`, `design`, `approach`, `requirements`, `clarify`.

---

## How to Use

### 1. Step-by-Step Instructions

1. **Explore Project Context:** Check codebase, existing architecture, and recent commits before proposing changes.
2. **Interactive Socratic Questioning:** Ask clarifying questions one at a time to uncover purpose, constraints, and success criteria.
3. **Structured Trade-Off Matrix:** Propose 2–3 approaches formatted as a Trade-off Matrix (Pros/Cons/Risk Class) with a recommended default choice.
4. **Domain A–G Evaluation:** Systematically evaluate design against Security, UI/UX, Performance, Reliability, Maintainability, Observability, and Compliance.
5. **High-Risk Class Grilling:** For Auth, Billing, Schema, API Gateway, or Secrets, execute One-Question Grilling and invoke `ag-docs adr` for hard architectural decisions.
6. **Write Design Doc / Formal Spec:**
   - For standard features, save design spec to `docs/design/<feature-topic>-design.md` using template `references/design-spec-template.md`.
   - For **High-Risk features** (Auth, Billing, DB Schema, API Gateway, Secrets), author the **Formal Specification** at `process/features/{feature}/active/{feature-slug}-{topic-slug}-formal-spec.md` using template `process/development-protocols/references/formal-spec-template.md` (defining Zod validation contracts, System Invariants `INV-1..N`, and Adversarial Payloads `ADV-1..N`).
7. **Transition:** After user approves written spec, invoke the `ag-generate-plan` skill to create the implementation plan.

### High-Risk Features & Formal Specifications (Architect Phase Gate)

When brainstorming work falls under **High-Risk Class** (Auth, Billing & Payments, Database Schema Migrations, Public API Contracts, Gateway, or Secrets Management):

1. **Invoke Architect / Verifier Protocol:** Execute One-Question Grilling per `process/development-protocols/references/architect-verifier-master-workflow-guide.md`.
2. **Author Formal Specification:** Create the Formal Spec artifact at `process/features/{feature}/active/{feature-slug}-{topic-slug}-formal-spec.md` using template `process/development-protocols/references/formal-spec-template.md`.
3. **Mandatory Formal Spec Elements:**
   - **Strict Input Data Schema:** Define Zod schemas or TypeScript Discriminated Unions for all inputs.
   - **System Invariants (`INV-1..N`):** State mathematical or logical rules that MUST ALWAYS HOLD TRUE.
   - **Adversarial Validation Matrix (`ADV-1..N`):** Define Level 2 edge cases (concurrency, race conditions, malicious payloads) and fail-safe recovery behaviors.
4. **ADR Creation:** Record any hard architectural decisions or trade-offs via `ag-docs adr` at `docs/adr/000X-<kebab-case-name>.md`.

### 2. Examples

#### Example 1: New Feature Brainstorming & Trade-Off Matrix

- **Context:** User wants to add OAuth2 social login to an existing app.
- **User prompt:** "Help me brainstorm the design for adding Google and GitHub OAuth login."
- **Action/Result:** Agent checks current auth state, presents a Trade-off Matrix comparing Session cookies vs JWT tokens, evaluates Domain A (Security) and Domain D (Reliability), writes spec doc, and transitions to plan mode.

#### Example 2: High-Risk System Redesign

- **Context:** Redesigning payment credit calculation and schema migration.
- **User prompt:** "Brainstorm redesigning our billing credit engine."
- **Action/Result:** Agent identifies High-Risk billing class, executes One-Question Grilling with default choices, records ADR via `ag-docs adr`, writes Formal Spec to `process/features/billing/active/Billing_Credit_Formal_Spec.md`, and requests user review.

---

## Quick Reference

- `collaborative-design` - Never write code before design approval.
- `trade-off-matrix` - Present choices using a structured Pros/Cons/Risk matrix.
- `domain-a-g-framework` - Evaluate across Security, UI/UX, Performance, Reliability, Maintainability, Observability, and Compliance.
- `high-risk-grilling` - Enforce system invariants (`INV-1`) and ADR creation (`ag-docs adr`) for critical domains.
- `terminal-state` - The only skill invoked after design approval is `ag-generate-plan`.

---

## References

- [references/brainstorming-guide.md](references/brainstorming-guide.md) - Complete workflow diagram, Trade-off Matrix template, 7 Engineering Evaluation Domains (A–G), and 8-step checklist.
- [references/design-spec-template.md](references/design-spec-template.md) - Standard Design Specification template for non-high-risk features.
- [process/development-protocols/references/formal-spec-template.md](../../../process/development-protocols/references/formal-spec-template.md) - Formal Specification template for High-Risk features.
- [process/context/all-context.md](../../../process/context/all-context.md) - Agent Skills Kit Repository Context and codebase guidelines.

From ngxccc/agent-skills-kit, MIT.

More from ngxccc/agent-skills-kit

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Aug 2, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from ngxccc

Cite this page

ag-brainstorming skill. Agent Mart. https://agentmart.dev/skills/ngxccc-ag-brainstorming (last updated 2026-10-24)