AI Skill MartBrowse skillsPromptsProgramsLearnWhat is a Skill?
 
← Back to list

EGM Brainstorm (Multi-Lens Ideation & Refinement)

Turns half-baked feature ideas into a stress-tested RFC brief through multi-turn questioning, six parallel critique lenses, and an iterative refinement loop.

Dev & CodingIntermediate★ 6,600⑂ 402AI score 8/10Last updated: Oct 30, 2026

What it does

  • Drills into a raw concept with open-ended questions across multiple turns — problem statement, consumers/SLAs, hard constraints, out-of-scope edges, success metrics.
  • Enforces a readiness gate: no reviewers are spawned until you approve a 2–3 sentence anchored hypothesis.
  • Spawns the egm-synthesizer subagent, which runs six critique lenses in parallel: Contrarian/YAGNI, cost & token economics, UX & developer ergonomics, migration & backward compatibility, observability/SRE, long-term tech debt.
  • Artifact-first: findings are deduplicated and written to concepts_brief.md (Overview, Analysis, Goals vs Non-Goals, Alternatives Weighed, Key Decisions) rather than dumped into chat.
  • Ends with a handoff gate: publish as an RFC under docs/designs/, hand off to /egm-draft, keep iterating, or park the brief.
  • Hard behavioral rules: no sycophancy, critique first, generate counter-arguments, no reflexive apologies.

Who it's for

  • Engineers and tech leads who want holes poked in an idea before writing a design doc.
  • Small teams without enough reviewers for a real design review.
  • Anyone who wants an adversarial reviewer rather than an agreeable assistant.

Example uses

  1. "Add LLM-based log summarization" — questioning pins down peak QPS, token budget and false-positive tolerance; the cost lens flags quota blowup at scale.
  2. "Add a new CLI subcommand" — the YAGNI lens asks whether an existing library covers 90% of it; the UX lens checks whether errors are actionable.
  3. Publish the finished brief to docs/designs/<feature>.md for team feedback, then run /egm-draft docs/designs/<feature>.md to produce the formal design document.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/google/elephant-goldfish-model/HEAD/distributions/claude/skills/egm-brainstorm/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on AI Skill Mart.
Copy the distributions/claude/skills/egm-brainstorm folder from the GitHub repo google/elephant-goldfish-model into my ~/.claude/skills/egm-brainstorm/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/google/elephant-goldfish-model.git && cp -r elephant-goldfish-model/distributions/claude/skills/egm-brainstorm ~/.claude/skills/egm-brainstorm

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal in the directory where you keep tooling.
  2. Clone the repo: git clone https://github.com/google/elephant-goldfish-model.git
  3. Copy the skill: cp -r elephant-goldfish-model/distributions/claude/skills/egm-brainstorm ~/.claude/skills/egm-brainstorm
  4. The skill references sibling files (agents/synthesizer.md, docs/claude.md, docs/harnesses.md), so install the whole distributions/claude bundle as a plugin or copy those files along with it.
  5. Restart Claude Code and invoke it explicitly with /egm-brainstorm — model auto-invocation is disabled in its frontmatter.
  6. Make sure your harness has a subagent/Agent tool; without it the skill stops and tells you a same-context review is not a valid Goldfish review.
View source on GitHub ↗License: Apache-2.0