home / skills / nickcrew / claude-cortex / idea_lab

idea_lab skill

safe

This skill helps generate rapid divergent ideas with 5–7 concepts, selects top 3, and assigns 1-day experiments for fast validation.

npx playbooks add skill nickcrew/claude-cortex --skill idea_lab

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

1.1 KB

---
name: idea-lab
description: Timeboxed divergent ideation that outputs ranked options plus day-one experiments.
license: MIT
---

# `/collaboration:idea-lab`

Use this when you need many angles fast and want to leave with 1-day experiments queued.

## Inputs
- Topic or problem statement
- Constraints (time, platform, compliance, budget)
- Any existing assets to reuse

## Steps
1. State the timebox (default 15m).\n2. Capture goal, success signals, constraints, assets.\n3. Generate 5–7 distinct concepts with Wow-Factor, Feasibility (S/M/L), Dependency to check.\n4. Select Top 3 to test today; assign a 1-day experiment to each.\n5. Seed Task view or hand off to `/ctx:plan`.

## Output Template
```
### Problem / Goal
### Success Signals
### Constraints
### Existing Assets
### Options (table)
| Concept | Wow-Factor | Feasibility | Dependency to Check |
|---------|------------|-------------|---------------------|
### Top 3 Experiments
### Next Steps
```

## Resources
- See `modes/Idea_Lab.md` for tone cues.\n- Pair with `/ctx:plan` after choosing the top experiment.\n- Use `/collaboration:assumption-buster` first if the problem is sticky.

Overview

This skill runs timeboxed divergent ideation sessions that produce ranked options plus actionable day-one experiments. It quickly converts broad problems into a short list of testable concepts and immediate next steps. The goal is fast alignment and momentum: leave the session with three experiments you can start within a day.

How this skill works

Set a timebox (default 15 minutes) and capture the goal, success signals, constraints, and any reusable assets. The skill generates 5–7 distinct concepts evaluated by Wow-Factor, Feasibility (S/M/L), and Dependencies to check. It then selects the top three concepts and assigns a one-day experiment to each, seeding follow-up work into a planning context or handoff.

When to use it

When you need many distinct approaches fast and must pick experiments to run immediately
During sprint planning to seed rapid validation work
When a stakeholder asks for creative options under tight constraints
Before committing engineering or design resources to a single direction
When you want to convert brainstorming into measurable, one-day tests

Best practices

Keep the timebox strict (10–20 minutes) to encourage divergent thinking and avoid overanalysis
Define clear success signals up front so experiments are measurable
Capture constraints and existing assets to focus ideas on realistic options
Force diversity: aim for 5–7 concepts across different approaches or tech stacks
Assign clear owners and one-day scopes for each experiment to ensure immediate progress

Example use cases

Product team exploring three rapid prototypes to increase onboarding conversion
Growth team testing three acquisition channel experiments before the end of the week
Designers generating interface alternatives with quick usability checks
Engineering evaluating three low-effort feature toggles to measure impact
Enterprise stakeholders aligning on compliant options under budget and platform limits

FAQ

How long should the timebox be?

Default is 15 minutes; use 10–20 minutes depending on complexity and participant familiarity.

What does a one-day experiment look like?

A scoped, owner-assigned task with clear success signals and minimal dependencies you can complete or validate within a single workday.

What if the problem has many unknowns?

Run an assumption-buster session first to surface and prioritize key unknowns, then use the idea lab to generate experiments targeting those risks.