Agentic Engineering Playbook
PM Kit

Ideation — Brainstorming Lab

Select the right divergence strategy for your problem before committing to any concept. The Brainstorming Lab acts as an entry point and routes your session to one of 20 curated strategies.

Phase 0 — Ideation

Ideation is the work that happens before the charter. You don't yet have an approved project name, a sponsor signing paperwork, or a budget. What you have is a situation: an unsolved problem, an opportunity to explore, or a question the team hasn't been able to answer. The ideation phase converts that vague situation into a concept clear enough to start a project on solid footing.

In the project management lifecycle, ideation feeds directly into the initiation phase. A well-structured concept makes the charter easier to draft, the stakeholder register more accurate, and the product brief more honest. If you skip ideation or do it superficially, every phase that follows pays the cost: the team re-debates decisions that should already be made.

When this phase begins

The ideation phase begins when someone — a sponsor, a teacher, a team — identifies a need or opportunity without a formal project existing yet. It ends when the team produces a concept memo that the sponsor can read and, if convinced, use as input for the charter.


Artifact: Brainstorming Lab

Skillbrainstorming-lab
TypeRouter / meta-skill
When to useThe team wants structured brainstorming but doesn't know which technique to use. The Lab asks diagnostic questions and recommends one or two strategies from the curated deck of 20.
Acceptance checklistdocs/pm-kit/checklists/brainstorming-lab.md inside your installed project

Invocation pattern. Open your agent (Claude Code or Gemini CLI) in your project directory and instruct:

Invoke the brainstorming-lab skill.

The agent first asks you to name the goal of your session. If your goal matches its shortcut table, the recommendation is immediate. If not, it falls back to up to three diagnostic questions covering problem type, team size, and available time, then recommends the most appropriate strategy. Once the selection is ready, the agent saves the rationale to docs/pm-kit/outputs/brainstorming-lab/.


How to pick the right strategy

The Brainstorming Lab is the correct entry point when you're not sure which technique to apply. But if you already know your situation, you can invoke any of the 20 strategies directly.

Start by answering these two questions:

  1. Do you need to expand the solution space (generate more ideas) or narrow it (find the best path)?
  2. Where are you in the project: early discovery, a specific blocker, or a decision among already-known options?

If you don't have clear answers, invoke brainstorming-lab and let the agent run the diagnosis.


The 20 strategies in the curated deck

Each strategy is a standalone skill with a facilitation script included. Invoke whichever one the Lab recommends, or choose directly if you already know the technique you need.

Show all 20 strategies
SkillWhen to use
brainstorming-five-whysRoot-cause analysis; best for convergent problem diagnosis
brainstorming-question-stormingGenerate questions before answers; best for early discovery
brainstorming-assumption-reversalFlip core assumptions; best for paradigm-shift moments
brainstorming-constraint-mappingSurface and challenge limitations; best for blocked planning
brainstorming-failure-analysisLearn from failures; best for risk identification
brainstorming-morphological-analysisSystematic parameter combinations; best for complex option spaces
brainstorming-six-thinking-hatsMulti-perspective parallel thinking; best for teams needing structured debate
brainstorming-mind-mappingVisual idea branching; best for solo or small-team exploration
brainstorming-decision-tree-mappingMap decision paths and outcomes; best for choice evaluation
brainstorming-solution-matrixGrid of variables and approaches; best for systematic option comparison
brainstorming-resource-constraintsExtreme limitation imposition; best for forcing creative priorities
brainstorming-role-playingStakeholder perspective embodiment; best for empathy and stakeholder discovery
brainstorming-brain-writing-round-robinSilent written idea building in rotation; best for inclusive large groups
brainstorming-reverse-brainstormingGenerate problems to reveal solutions; best for risk and opportunity discovery
brainstorming-what-if-scenariosRadical possibility exploration; best for breaking stuck thinking
brainstorming-first-principles-thinkingRebuild from fundamental truths; best for breakthrough innovation
brainstorming-values-archaeologyExcavate deep motivating values; best for alignment and priority conflicts
brainstorming-alien-anthropologistOutsider's bewildered perspective; best for surfacing hidden assumptions
brainstorming-chaos-engineeringDeliberate stress-testing; best for resilience and risk planning
brainstorming-anti-solutionGenerate ways to make the problem worse; best for revealing hidden assumptions

What ideation must get right

  • Don't skip the diagnosis. Choosing the wrong strategy produces ideas that don't fit the problem type. The Lab's short diagnosis takes minutes and saves hours later.
  • Capture the reasoning, not just the ideas. The concept memo the Lab produces must explain why the team chose that angle, not just what ideas emerged. The charter will need it.
  • One approved concept, not ten. Ideation ends when there is a single concept clear enough for the sponsor to authorize. Two or three candidates without a decision is not an ideation deliverable — it's unfinished planning work.

On this page