Skip to content
Agentmart
Submit a tool

hm-l2-brainstorm

Use when the user says "let's brainstorm", "help me think through", "figure out what to build", "clarify requirements", "explore ideas", "what should we…

Skill by shynlee04

IndexedMIT★ 51
Write a review

Chalk Labs

Need help deploying hm-l2-brainstorm?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

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/shynlee04-hm-l2-brainstorm

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 the user says "let's brainstorm", "help me think through", "figure out what to build", "clarify requirements", "explore ideas", "what should we build", "ideation", "requirements gathering", "I have a vague idea", or when user intent is ambiguous and needs structured exploration before spec-driven-authoring or test-driven-execution can begin. This skill bridges vague intent to a formal requirements brief ready for handoff to hm-spec-driven-authoring. NOT for solution design, architecture decisions, or implementation planning — those come AFTER requirements are surfaced.

Use cases: design, testing

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 and code. Checked Nov 1, 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/hm-l2-brainstorm

What's inside

Skill name
hm-l2-brainstorm
Description
Use when the user says "let's brainstorm", "help me think through", "figure out what to build", "clarify requirements", "explore ideas", "what should we build", "ideation", "requirements gathering", "I have a vague idea", or when user intent is ambiguous and needs structured exploration before spec-driven-authoring or test-driven-execution can begin. This skill bridges vague intent to a formal requirements brief ready for handoff to hm-spec-driven-authoring. NOT for solution design, architecture decisions, or implementation planning — those come AFTER requirements are surfaced.
License
MIT
Metadata
role: domain-execution · layer: 2 · pattern: P2 · version: 1.0.0
Instructions
297 lines in SKILL.md · view on GitHub

Agent Skills spec issues

  • allowed-tools: must be a space-separated string

Files (8)

  • SKILL.md
  • evals.json · evals/
  • rich-gate-scorecard.md · metrics/
  • assumption-detection.md · references/
  • handoff-template.md · references/
  • question-patterns.md · references/
  • .gitkeep · scripts/
  • task_plan.md

SKILL.md (first lines)

# Brainstorming: Intent → Requirements

Bridge vague user intent to a formal requirements brief through structured questioning, assumption detection, and scope surfacing. This skill owns the ideation-to-requirements phase. It does NOT design solutions — only clarifies WHAT needs to be built.

This package synthesizes three inspected third-party patterns:

| Source | Adopt / Adapt decision | Local transformation |
|--------|------------------------|----------------------|
| `obra/superpowers@brainstorming` | Adopt sequential questioning, hard-gate-before-code, one-question-at-a-time, approach comparisons. Adapt away from superpowers-specific paths. | Use Gate 1 (no implementation before requirements brief approved), one-question-at-a-time, approach comparison in Phase 2. |
| `ratacat/claude-skills@brainstorming` | Adopt YAGNI principles, question techniques, anti-patterns table, handoff options. Adapt generic design doc → hm-specific requirements brief. | Use YAGNI filter in Phase 2, question techniques in Phase 1, anti-patterns throughout. |
| `Jamie-BitFlight/brainstorming-skill` | Adopt assumption challenge technique, constraint variation. Defer creative ideation patterns (perspective multiplication, inversion, SCAMPER). | Use assumption challenge in Phase 1.3, constraint variation in Phase 2.1. |

## HARD GATE — No Code Before Requirements Brief

```
DO NOT write code, scaffold projects, or invoke implementation skills
until the requirements brief is produced AND approved by the user.
This applies to EVERY project regardless of perceived simplicity.
```

## Entry Gate

Proceed when any of these are true:
- User said "brainstorm", "ideation", "figure out what to build", "explore ideas"
- User intent is vague or underspecified (missing constraints, success criteria, scope)
- User describes a problem without specifying a solution direction
- A coordinator skill (`hm-user-intent-interactive-loop`, `hm-coordinating-loop`) routes to this skill

Do NOT proceed when:
- A PRD, spec, or formal requirements document already exists → route to `hm-spec-driven-authoring`
- The task is a straightforward bug fix with clear expected behavior → execute directly
- The user says "just do it", "I know exactly what I want" with explicit details

## Checklist

Create a task for each item and complete them in order. Do not skip phases.

- [ ] Phase 1: Understand Intent — structured questioning, assumption detection, scope surfacing
- [ ] Phase 2: Explore Requirements — approach options, constraint mapping, priority setting
- [ ] Phase 3: Produce Requirements Brief — structured output for handoff
- [ ] Phase 4: Handoff — route to next skill (hm-spec-driven-authoring, hm-deep-research, or hm-detective)

## Phase 1: Understand Intent

### 1.1 Gather Context

Before asking the user any questions:
1. Check project files: `README.md`, instruction files (`AGENTS.md` or equivalent), `package.json`, recent commits
2. Identify the project domain, stack, and existing patterns
3. Check for existing specs, PRDs, or design docs that relate to the request
4. If the request mentions a specific file or module, read it

### 1.2 Assess Scope

Flag scope issues immediately — do NOT burn questions on something that needs decomposition first.

**Multi-system signal:** If the request describes multiple independent subsystems (e.g., "build a platform with auth, payments, and real-time chat"), stop and say:

> "This request spans multiple independent systems: [list them]. Each needs its own requirements brief. Which should we start with?"

**Appropriately scoped:** A single feature, component, module, or workflow. Continue to Phase 1.3.

### 1.3 Ask Clarifying Questions

**One question at a time.** Prefer multiple choice when natural options exist. Break complex topics into multiple questions.

Use question patterns from `references/question-patterns.md`. Default priority order:

| Priority | Category | Example |
|----------|----------|---------|
| 1 | Purpose | "What problem does this solve? (a) automate existing workflow, (b) add new capability, (c) replace legacy system" |
| 2 | Users | "Who uses this? (a) developers, (b) end-users, (c) admins — or multiple?" |
| 3 | Constraints | "Any hard constraints? Technology, timeline, compliance, dependencies?" |
| 4 | Success | "How will you know this is working? What does success look like?" |
| 5 | Scope boundaries | "What should this explicitly NOT do?" |

**Stop conditions (any of these):**
- User says "proceed", "let's move on", "that's enough questions"
- 5 questions asked (stop and surface what you have)
- All 5 priority categories covered with clear answers
- The intent is clear enough to state: "We need to build [X] so that [Y] can [Z]"

### 1.4 Detect Unstated Assumptions

After gathering answers, run assumption detection using `references/assumption-detection.md`:

1. List every assumption you made about the user's request
2. Validate explicit assumptions with the user: "I'm assuming [X]. Is that correct?"
3. Surface implicit assumptions the user may not have considered: "This implies [Y] will need to happen. Is that in scope?"

Record all validated assumptions in the requirements brief.

### 1.5 Pause for Confirmation

After questions are exhausted (max 5), restate your understanding:

```markdown
## Intent Summary

**Problem:** [1-2 sentences]
**Target users:** [who]
**Success looks like:** [measurable outcome]
**In scope:** [what's included]
**Out of scope:** [what's explicitly excluded]
**Key constraints:** [hard boundaries]
**Validated assumptions:** [confirmed assumptions]

Does this match what you had in mind? Adjustments before we explore requirements?
```

Wait for user confirmation before proceeding to Phase 2.

## Phase 2: Explore Requirements

### 2.1 Propose 2-3 Approach Directions

Not implementations — requirement shapes. For each approach, describe:

```markdown

From shynlee04/hivemind-plugin, MIT.

More from shynlee04/hivemind-plugin

Signals

Stars
51
Forks
16
Open issues
21
Last commit (this folder)
May 7, 2026 (6 months ago)

From GitHub (fetched Oct 24, 2026).

Details

Version
1.0.0
License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from shynlee04

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get hm-l2-brainstorm running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying hm-l2-brainstorm?

CallGet help
Cite this page

hm-l2-brainstorm skill. Agent Mart. https://agentmart.dev/skills/shynlee04-hm-l2-brainstorm (last updated 2026-10-24)